draft-ietf-oauth-v2-threatmodel.xml

<?xml version="1.0" encoding="US-ASCII"?>
<!DOCTYPE rfc SYSTEM "rfc2629.dtd">
<?rfc toc="yes"?>
<?rfc tocompact="yes"?>
<?rfc tocdepth="4"?>
<?rfc tocindent="yes"?>
<?rfc symrefs="yes"?>
<?rfc sortrefs="yes"?>
<?rfc comments="yes"?>
<?rfc inline="yes"?>
<?rfc compact="yes"?>
<?rfc subcompact="no"?>
<?rfc strict="no"?>
<rfc category="std" docName="draft-ietf-oauth-v2-threatmodel-00"
     ipr="trust200902">
  <front>
    <title abbrev="OAuth 2.0 Security">OAuth 2.0 Threat Model and Security
    Considerations</title>

    <author fullname="Torsten Lodderstedt" initials="T." role="editor"
            surname="Lodderstedt">
      <organization>Deutsche Telekom AG</organization>

      <address>
        <email>torsten@lodderstedt.net</email>
      </address>
    </author>

    <author fullname="Mark McGloin" initials="M." surname="McGloin">
      <organization>IBM</organization>

      <address>
        <email>mark.mcgloin@ie.ibm.com</email>
      </address>
    </author>

    <author fullname="Phil Hunt" initials="P." surname="Hunt">
      <organization>Oracle Corporation</organization>

      <address>
        <email>phil.hunt@yahoo.com</email>
      </address>
    </author>

    <date day="05" month="August" year="2011" />

    <area>Security Area</area>

    <workgroup>Web Authorization Protocol (oauth)</workgroup>

    <keyword>security</keyword>

    <keyword>oauth 2.0</keyword>

    <keyword>threat model</keyword>

    <abstract>
      <t>This document gives security considerations based on a comprehensive
      threat model for the OAuth 2.0 Protocol.</t>
    </abstract>

    <note title="Requirements Language">
      <t>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
      "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
      document are to be interpreted as described in <xref
      target="RFC2119">RFC 2119</xref>.</t>
    </note>
  </front>

  <middle>
    <section title="Introduction">
      <t>This document gives security considerations based on a comprehensive
      threat model for the OAuth 2.0 Protocol <xref
      target="I-D.ietf-oauth-v2"></xref>. It contains the following
      content:<list style="symbols">
          <t>Documents any assumptions and scope considered when creating the
          threat model.</t>

          <t>Describes the security features in-built into the OAuth protocol
          and how they are intended to thwart attacks.</t>

          <t>Gives a comprehensive threat model for OAuth and describes the
          respective counter measures to thwart those threats.</t>
        </list>Threats include any intentional attacks on OAuth tokens and
      resources protected by OAuth tokens as well as security risks introduced
      if the proper security measures are not put in place. Threats are
      structured along the lines of the protocol structure to aid development
      teams implement each part of the protocol securely. For example all
      threats for granting access or all threats for a particular client
      profile or all threats for protecting the resource server.</t>
    </section>

    <section title="Overview">
      <t></t>

      <section title="Scope">
        <t>The security considerations document only considers clients bound
        to a particular deployment as supported by <xref
        target="I-D.ietf-oauth-v2"></xref>. Such deployments have the
        following characteristics:</t>

        <t><list style="symbols">
            <t>Resource server URLs are static and well-known at development
            time, authorization server URLs can be static or discovered.</t>

            <t>Token scope values (e.g. applicable URLs and methods) are
            well-known at development time.</t>

            <t>Client registration: Since registration of clients is out of
            scope of the current core spec, this document assumes a broad
            variety of options from static registration during development
            time to dynamic registration at runtime.</t>
          </list>The following are considered out of scope :</t>

        <t><list style="symbols">
            <t>Communication between authorization server and resource
            server</t>

            <t>Token formats</t>

            <t>Except for &bdquo;Resource Owner Password Credentials&ldquo;
            (see <xref target="I-D.ietf-oauth-v2"></xref>, section 4.3), the
            mechanism used by authorization servers to authenticate the
            user</t>

            <t>Mechanism by which a user obtained an assertion and any
            resulting attacks mounted as a result of the assertion being
            false.</t>

            <t>Clients are not bound to a specific deployment: An example
            could by a mail client with support for contact list access via
            the portable contacts API (see <xref
            target="portable-contacts"></xref>). Such clients cannot be
            registered upfront with a particular deployment and must
            dynamically discover the URLs relevant for the OAuth protocol.</t>
          </list></t>
      </section>

      <section title="Attack Assumptions">
        <t>The following assumptions relate to an attacker and resources
        available to an attacker:</t>

        <t><list style="symbols">
            <t>It is assumed the attacker has full access to the network
            between the client and authorization servers and the client and
            the resource server, respectively. The attacker may eaves drop on
            any communications between those parties. He is not assumed to
            have access to communication between authorization and resource
            server.</t>

            <t>It is assumed an attacker has unlimited resources to mount an
            attack.</t>

            <t>It is assumed that 2 of the 3 parties involved in the OAuth
            protocol may collude to mount an attack against the 3rd party. For
            example, the client and authorization server may be under control
            of an attacker and collude to trick a user to gain access to
            resources.</t>
          </list></t>
      </section>

      <section title="Architectural assumptions">
        <t>This section documents the assumptions about the features,
        limitations and design options of the different entities of a OAuth
        deployment along with the security-sensitive data-elements managed by
        those entity. These assumptions are the foundation of the treat
        analysis.</t>

        <t>The OAuth protocol leaves deployments with a certain degree of
        freedom how to implement and apply the standard. The core
        specification defines the core concepts of an authorization server and
        a resource server. Both servers can be implemented in the same server
        entity, or they may also be different entities. The later is typically
        the case for multi-service providers with a single authentication and
        authorization system, and are more typical in middleware
        architectures.</t>

        <section title="Authorization Servers">
          <t>The following data elements MAY be stored or accessible on the
          authorization server:</t>

          <t><list style="symbols">
              <t>user names and passwords</t>

              <t>client ids and secrets</t>

              <t>client-specific refresh tokens</t>

              <t>client-specific access tokens (in case of handle-based
              design)</t>

              <t>HTTPS certificate/key</t>

              <t>per authorization process (in case of handle-based design):
              redirect_uri, client_id, authorization code</t>
            </list></t>
        </section>

        <section title="Resource Server">
          <t>The following data elements MAY be stored or accessible on the
          resource server:</t>

          <t><list style="symbols">
              <t>user data (out of scope)</t>

              <t>HTTPS certificate/key</t>

              <t>authz server credentials (handle-based design), or</t>

              <t>authz server shared secret/public key (assertion-based
              design)</t>

              <t>access tokens (per request)</t>
            </list> It is assumed that a resource server has no knowledge of
          refresh tokens, user passwords, or client secrets.</t>
        </section>

        <section title="Client">
          <t>The following data elements are stored or accessible on the
          client:</t>

          <t><list style="symbols">
              <t>client id (and client secret or corresponding client
              credential)</t>

              <t>one or more refresh tokens (persistent) and access tokens
              (transient) per end-user or other security-context or delegation
              context</t>

              <t>trusted CA certificates (HTTPS)</t>

              <t>per authorization process: redirect_uri, authorization
              code</t>
            </list></t>

          <section title="Web Application">
            <t>A web application is a client running on a web server,
            typically with its own user management. End-users access the
            client via an HTML user interface rendered in a user- agent on the
            end-user's device. The client credentials as well as any token
            issued to the client are stored on the web server and are not
            exposed to or accessible by the end-user. Tokens are bound to a
            single user identity at the site. The potential number of tokens
            affected by a security breach depends on number of site users.</t>

            <t>Such clients are implemented using the authorization code grant
            type (see <xref target="code_flow"></xref>).</t>
          </section>

          <section title="Native Applications">
            <t>A native application is a client which is installed and
            executes on the end-user's device, such as a notebook, PC, Tablet,
            Smartphone, or Gaming Console. The OAuth protocol data and
            credentials are accessible to the end-user. It is assumed that
            such an application can protect dynamically issued credentials,
            such as refresh tokens, from eavesdropping by other applications
            residing on the same device.</t>

            <t>Massively distributed applications such as these cannot
            reliably keep secrets confidential, which are issued per software
            package. This is because such secrets would need to be transferred
            to the user device as part of the installation process. An
            attacker could reverse engineer any secret from the binary or
            accompanying resources. Native Applications are able to protect
            per installation/instance secrets (e.g. refresh tokens) to some
            extent.</t>

            <t>Device platforms typically allow users to lock the device with
            a PIN code and to segregate different applications or users
            (multi-user operation systems).</t>

            <t>Some devices can be identified/authenticated (to varying
            degrees of assurance):</t>

            <t><list style="symbols">
                <t>Handsets and smart phones by its International Mobile
                Equipment Identity (IMEI)</t>

                <t>Set top boxes, gaming consoles, others by using
                certificates and TPM module - Note: This does not help to
                identify client applications but may be used to bound tokens
                to devices and to detect token theft</t>
              </list>Mobile devices, such as handsets or smart phones have the
            following special characteristics:</t>

            <t><list style="symbols">
                <t>Limited input capabilities, therefore such clients
                typically obtain a refresh token in order to provide automatic
                login for sub-sequent application sessions</t>

                <t>As mobile and small devices, they can get cloned, stolen or
                lost easier than other devices.</t>

                <t>Security breach will affect single user (or a few users)
                only.</t>
              </list>For the purposes of this document, the scenario of
            attackers who control a smartphone device entirely is out of
            scope.</t>

            <t>There are several implementation options for native
            applications:</t>

            <t><list style="symbols">
                <t>The authorization code grant type in combination with an
                embedded or external browser (<xref
                target="code_flow"></xref>)</t>

                <t>The implicit grant type in combination with an embedded or
                external browser (<xref target="implicite_flow"></xref>)</t>

                <t>The resource owner password credentials grant type can be
                used as well (<xref target="pwd_flow"></xref>)</t>
              </list>Different threats exists for those implementation
            options, which are discussed in the respective sections of the
            threat model.</t>
          </section>

          <section title="User-agent-based Applications">
            <t>A user-agent-based application is a client in which the client
            code is downloaded from a web server and executes within a
            user-agent on the end-user's device. The OAuth protocol data and
            credentials are accessible to the end-user. Since such
            applications directly reside within the user-agent, they can make
            seamless use of the user-agent capabilities in the end-user
            authorization process.</t>

            <t>Such client are implemented using the implicit grant type
            (<xref target="implicite_flow"></xref>).</t>
          </section>

          <section title="Autonomous">
            <t>Autonomous clients access resource services using rights grants
            by client credentials only. Thus the autonomous client becomes the
            &bdquo;user&ldquo;. Authenticating autonomous clients is
            conceptually similar to end-user authentication since the issued
            tokens refer to the client's identity. Autonomous clients shall
            always be required to use a secret or some other form of
            authentication (e.g. client assertion in the form of a SAML
            assertion or STS token) acceptable to the authorization/token
            services. The client must ensure the confidentiality of
            client_secret or other credential.</t>

            <t>Such client are implemented using the client credentials grant
            type.</t>
          </section>
        </section>
      </section>
    </section>

    <section title="Security Features">
      <t>These are some of the security features which have been built into
      the OAuth 2.0 protocol to mitigate attacks and security issues.</t>

      <section title="Tokens">
        <t>OAuth makes extensive use of all kinds of tokens (access tokens,
        refresh tokens, authorization codes). The information content of a
        token can be represented in two ways as follows:</t>

        <t><list style="hanging">
            <t hangText="Handle (or artifact)">a reference to some internal
            data structure within the authorization server, the internal data
            structure contains the attributes of the token, such as user id,
            scope, etc. Handles enable simple revocation and do not require
            cryptographic mechanisms to protected token content from being
            modified. On the other hand, handles require communication between
            issuing and consuming entity (e.g. authorization and resource
            server) in order to validate the token and obtain token-bound
            data. This communication might have an negative impact on
            performance and scalability if both entities reside on different
            system. Handles are therefore typically used if the issuing and
            consuming entity are the same. A 'handle' token is often referred
            to as an 'opaque' token because the resource server does not need
            to be able to interpret the token directly, it simply uses the
            token.</t>

            <t hangText="Assertions (aka self-contained token)">a parseable
            token. An assertion typically has a duration, an audience, and is
            digitally signed containing information about the user and the
            client. Examples of assertion formats are SAML assertions and
            Kerberos tickets. Assertions can typically directly be validated
            and used by a resource server without interactions with the
            authorization server. This results in better performance and
            scalability in deployment where issuing and consuming entity
            reside on different systems. Implementing token revocation is more
            difficult with assertions than with handles.</t>
          </list>Tokens can be used in two ways to invoke requests on resource
        servers as follows:</t>

        <t><list style="hanging">
            <t hangText="bearer token">A 'bearer token' is a token that can be
            used by any client who has received the token (e.g. <xref
            target="I-D.ietf-oauth-v2-bearer"></xref>). Because mere
            possession is enough to use the token it is important that
            communication between end-points be secured to ensure that only
            authorized end-points may capture the token. The bearer token is
            convenient to client applications as it does not require them to
            do anything to use them (such as a proof of identity). Bearer
            tokens have similar characteristics to web single-sign-on (SSO)
            cookies used in browsers.</t>

            <t hangText="proof token">A 'proof token' is a token that can only
            be used by a specific client. Each use of the token, requires the
            client to perform some action that proves that it is the
            authorized user of the token. Examples of this are MAC tokens,
            which require the client to digitally sign the resource request
            with a secret corresponding to the particular token send with the
            request (e.g.<xref
            target="I-D.ietf-oauth-v2-http-mac"></xref>).</t>
          </list></t>

        <section title="Scope">
          <t>A Scope represents the access authorization associated with a
          particular token with respect to resource servers, resources and
          methods on those resources. Scopes are the OAuth way to explicitly
          manage the power associated with an access token. A scope can be
          controlled by the authorization server and/or the end-user in order
          to limit access to resources for OAuth clients these parties deem
          less secure or trustworthy. Optionally, the client can request the
          scope to apply to the token but only for lesser scope than would
          otherwise be granted, e.g. to reduce the potential impact if this
          token is sent over non secure channels. A scope is typically
          complemented by a restriction on a token's lifetime.</t>
        </section>

        <section title="Expires_In">
          <t>Expires_In allows an authorization server (based on its policies
          or on behalf of the end-user) to limit the lifetime of the access
          token. This mechanisms can be used to issue short-living tokens to
          OAuth clients the authorization server deems less secure or where
          sending tokens over non secure channels.</t>
        </section>
      </section>

      <section title="Access Token">
        <t>An access token is used by a client to access a resource. Access
        tokens typically have short life-spans (minutes or hours) that cover
        typical session lifetimes. An access token may be refreshed through
        the use of a refresh token. The short lifespan of an access token in
        combination with the usage of refresh tokens enables the possibility
        of passive revocation of access authorization on the expiry of the
        current access token.</t>
      </section>

      <section title="Refresh Token">
        <t>A refresh token represents a long-lasting authorization of a
        certain client to access resources on behalf of a resource owner. Such
        tokens are exchanged between client and authorization server, only.
        Clients use this kind of token to obtain ("refresh") new access tokens
        used for resource server invocations.</t>

        <t>A refresh token, coupled with a short access token lifetime, can be
        used to grant longer access to resources without involving end user
        authorization. This offers an advantage where resource servers and
        authorization servers are not the same entity, e.g. in a distributed
        environment, as the refresh token must always be exchanged at the
        authorization server. The authorization server can revoke the refresh
        token at any time causing the granted access to be revoked once the
        current access token expires. Because of this, a short access token
        lifetime is important if timely revocation is a high priority.</t>

        <t>The refresh token is also a secret bound to the client identifier
        and <spanx style="emph">instance</spanx> which originally requested
        the authorization and representing the original resource owner grant.
        This is ensured by the authorization process as follows:</t>

        <t><list style="numbers">
            <t>The resource owner and user-agent safely deliver the
            authorization code to the client instance in first place.</t>

            <t>The client uses it immediately in secure transport-level
            communications to the authorization server and then securely
            stores the long-lived refresh token.</t>

            <t>The client always uses the refresh token in secure
            transport-level communications to the authorization server to get
            an access token (and optionally rollover the refresh token).</t>
          </list>So as long as the confidentiality of the particular token can
        be ensured by the client, a refresh tokens can also be used as an
        alternative mean to authenticate the client instance itself.</t>
      </section>

      <section title="Authorization Code">
        <t>An Authorization Code represents the intermediary result of a
        successful end-user authorization process and is used by the client to
        obtain access and refresh token. Authorization codes are sent to the
        client's redirect_uri instead of tokens for two purposes.</t>

        <t><list style="numbers">
            <t>Instead of (longer-lasting) tokens, the short-living
            authorization code is exposed to potential attackers via URI query
            parameters (HTTP referrer), browser cacher or log file
            entries.</t>

            <t>It is much simpler to authenticate clients during the direct
            request between client and authorization server than in the
            context of the indirect authorization request. The later would
            require digital signatures.</t>
          </list></t>
      </section>

      <section title="Redirect-URI">
        <t>A redirect-uri helps to identify clients and prevents phishing
        attacks from other clients attempting to trick the user into believing
        the phisher is the client. The value of the actual redirect_uri used
        in the authorization request has to be presented and is verified when
        an authorization code is exchanged for tokens. This helps to prevent
        attacks, where the authorization code is revealed through redirectors
        and counterfeit web application clients. Moreover, the authorization
        server may require clients to pre-register their redirect URIs and
        validate the redirect_uri in the authorization request in order to
        detect malicious clients.</t>
      </section>

      <section title="State parameter">
        <t>The state parameter is used to link requests and callbacks to
        prevent CSRF attacks where an attacker authorizes access to his own
        resources and then tricks a users into following a redirect with the
        attacker's token.</t>
      </section>

      <section title="Client Identity">
        <t>Authentication protocols have typically not taken into account the
        identity of the software component acting on behalf of the end-user.
        OAuth does this in order to increase the security level in delegated
        authorization scenarios and because the client will be able to act
        without the user's presence. Depending on the client type, the client
        identity can and should be authenticated (see below).</t>

        <t>OAuth uses the <spanx style="emph">client_id</spanx> (client
        identity) to collate associated request to the same originator, such
        as</t>

        <t><list style="symbols">
            <t>a particular end-user authorization process and the
            corresponding request on the tokens endpoint to exchange the
            authorization code for tokens or</t>

            <t>the initial authorization and issuance of a tokens by an
            end-user to a particular client and sub-sequent requests by this
            client to obtain tokens w/o user consent (automatic processing of
            repeated authorization)</t>
          </list>The client identity may also be used by the authorization
        server to display relevant registration information to a user when
        requesting consent for scope requested by a particular client. The
        client identity may be used to limit the number of request for a
        particular client or to charge the client per request. Client Identity
        may furthermore be useful to differentiate access by different
        clients, e.g. in server log files.</t>

        <t>The <spanx style="emph">client_secret</spanx> is used to verify the
        client identifier. The authorization server should only rely on this
        form of client authentication where these secrets can be deployed to
        the clients in a secure manner and the client is capable of keeping
        its secret confidential. Alternatively, the client identity can also
        be verified using the <spanx style="emph">redirect_uri</spanx> or by
        the <spanx style="emph">end-user</spanx>.</t>

        <t>Clients (and the trustworthiness of its identity) can be classified
        by using the following parameters:</t>

        <t><list style="symbols">
            <t>Deployment-specific or -independent client_id (Note: for native
            applications, every installation of a particular application on a
            certain device is considered a deployment.)</t>

            <t>Validated properties, such as application name or
            redirect_uri</t>

            <t>Client_secret available</t>
          </list>Typical client categories are:</t>

        <t><list style="hanging">
            <t
            hangText="Deployment-independent client_id with pre-registered redirect_uri and without client_secret">Such
            an identity is used by multiple installations of the same software
            package. The identity of such a client can only be validated with
            the help of the end-user. This is a viable option for native
            applications in order to identify the client for the purpose of
            displaying meta information about the client to the user and to
            differentiate clients in log files. Revocation of such an identity
            will affect ALL deployments of the respective software.</t>

            <t
            hangText="Deployment-independent client_id with pre-registered redirect_uri and with client_secret">This
            is an option for native applications only, since web application
            would require different redirect URIs. This category is not
            advisable because the client secret cannot be protected
            appropriately (see <xref target="ObtainClientSecrets"></xref>).
            Due to its security weaknesses, such client identities have the
            same trust level as deployment-independent clients without secret.
            Revocation will affect ALL deployments.</t>

            <t
            hangText="Deployment-specific client_id with pre-registered redirect_uri and with client_secret">The
            client registration process insures the validation of the client's
            properties, such as redirect_uri, website address, web site name,
            contacts. Such a client identity can be utilized for all relevant
            use cases cited above. This level can be achieved for web
            applications in combination with a manual or user-bound
            registration process. Achieving this level for native applications
            is much more difficult. Either the installation of the application
            is conducted by an administrator, who validates the clients
            authenticity, or the process from validating the application to
            the installation of the application on the device and the creation
            of the client credentials is controlled end-to-end by a single
            entity (e.g. application market provider). Revocation will affect
            a single deployment only.</t>

            <t
            hangText="Deployment-specific client_id with client_secret without validated properties">Such
            a client can be recognized by the authorization server in
            transactions with subsequent requests (e.g. authorization and
            token issuance, refresh token issuance and access token
            refreshment). The authorization server cannot assure any property
            of the client to end-users. Automatic processing of
            re-authorizations could be allowed as well. Such client
            credentials can be generated automatically without any validation
            of client properties, which makes it another option especially for
            native applications. Revocation will affect a single deployment
            only.</t>
          </list>Use of the client secret is considered a relatively weak form
        of credential for the client. Use of stronger mechanisms such as a
        client assertion (e.g. SAML), key cryptography, are preferred.</t>
      </section>
    </section>

    <section title="Security Threat Model">
      <t>This sections gives a comprehensive threat model of OAuth 2.0.
      Threats are grouped first by attacks directed against an OAuth
      component, which are client, authorization server, and resource server.
      Subsequently, they are grouped by flow, e.g. obtain token or access
      protected resources. Every countermeasure description refers to a
      detailed description in <xref
      target="security_considerations"></xref>.</t>

      <section title="Clients">
        <t>This section describes possible threats directed to OAuth
        clients.</t>

        <section anchor="ObtainClientSecrets"
                 title="Threat: Obtain Client Secrets">
          <t>The attacker could try to get access to the secret of a
          particular client in order to:</t>

          <t><list style="symbols">
              <t>replay its refresh tokens and authorization codes, or</t>

              <t>obtain tokens on behalf of the attacked client with the
              privileges of that client.</t>
            </list>The resulting impact would be:</t>

          <t><list style="symbols">
              <t>Client authentication of access to authorization server can
              be bypassed</t>

              <t>Stolen refresh tokens or authorization codes can be
              replayed</t>
            </list>Depending on the client category, there are the following
          approaches an attacker could utilize to obtain the client
          secret.</t>

          <t><spanx style="strong">Attack: Obtain Secret From Source Code or Binary.</spanx>
          This applies for all client profiles. For open source projects,
          secrets can be extracted directly from source code in their public
          repositories. Secrets can be extracted from application binaries
          just as easily when published source is not available to the
          attacker. Even if an application takes significant measures to
          obfuscate secrets in their application distribution one should
          consider that the secret can still be reverse-engineered by anyone
          with access to a complete functioning application bundle or
          binary.</t>

          <t><spanx>Countermeasures:</spanx></t>

          <t><list style="symbols">
              <t>Don't issue secrets to clients with inappropriate security
              policy - <xref target="dont_issue"></xref></t>

              <t>Clients without secret require user consent - <xref
              target="forced_user_consent"></xref></t>

              <t>Use deployment-specific client secrets - <xref
              target="depl_specific_secretes"></xref></t>

              <t>Client secret revocation - <xref
              target="client_secret_revocation"></xref></t>
            </list> <spanx xml:space="default"></spanx></t>

          <t><spanx style="strong" xml:space="default">Attack: Obtain a
          Deployment-Specific Secret.</spanx> An attacker may try to obtain
          the secret from a client installation, either from a web site (web
          server) or a particular devices (native application).</t>

          <t><spanx>Countermeasures:</spanx></t>

          <t><list style="symbols">
              <t>Web server: apply standard web server protection measures
              (for config files and databases) - <xref
              target="std_web"></xref></t>

              <t>Native applications: Store secrets in a secure local storage
              - <xref target="secure_storage"></xref></t>

              <t>Client secret revocation - <xref
              target="client_secret_revocation"></xref></t>
            </list></t>
        </section>

        <section title="Threat: Obtain Refresh Tokens">
          <t>Depending on the client type, there are different ways refresh
          tokens may be revealed to an attacker. The following sub-sections
          give a more detailed description of the different attacks with
          respect to different client types and further specialized
          countermeasures. Some generally applicable countermeasure to
          mitigate such attacks shall be given in advance:</t>

          <t><list style="symbols">
              <t>The authorization server must validate the client id
              associated with the particular refresh token with every refresh
              request- <xref target="binding_refresh_client_id"></xref></t>

              <t>Limited scope tokens - <xref target="limit_scope"></xref></t>

              <t>Refresh token revocation - <xref
              target="refresh_revocation"></xref></t>

              <t>Client secret revocation - <xref
              target="client_secret_revocation"></xref></t>

              <t>Refresh tokens can automatically be replaced in order to
              detect unauthorized token usage by another party (Refresh Token
              Replacement) - <xref target="refresh_replace"></xref></t>
            </list> <spanx style="strong"></spanx></t>

          <t><spanx style="strong">Attack: Obtain Refresh Token from Web application.</spanx>
          An attack may obtain the refresh tokens issued to a web server
          client. Impact: Exposure of all refresh tokens on that side.</t>

          <t><spanx style="emph">Countermeasures:</spanx> <list
              style="symbols">
              <t>Standard web server protection measures - <xref
              target="std_web"></xref></t>

              <t>Use strong client authentication (e.g. client_assertion /
              client_token), so the attacker cannot obtain the client secret
              required to exchange the tokens - <xref
              target="strong_client_authn"></xref></t>
            </list> <spanx style="strong"></spanx></t>

          <t><spanx style="strong">Attack: Obtain Refresh Token from Native clients.</spanx>
          On native clients, leakage of a refresh token typically affects a
          single user, only.</t>

          <t><spanx style="emph">Read from local filesystem:</spanx> The
          attacker could try get file system access on the device and read the
          refresh tokens. The attacker could utilize a malicious application
          for that purpose.</t>

          <t><spanx style="emph">Countermeasures:</spanx></t>

          <t><list style="symbols">
              <t>Store secrets in a secure storage - <xref
              target="secure_storage"></xref></t>

              <t>Utilize device lock to prevent unauthorized device access -
              <xref target="device_lock"></xref></t>
            </list> <spanx style="emph"></spanx></t>

          <t><spanx style="emph">Steal device</spanx>: The host device (e.g.
          mobile phone) may be stolen. In that case, the attacker gets access
          to all applications under the identity of the legitimate user.</t>

          <t><spanx style="emph">Countermeasures:</spanx></t>

          <t><list style="symbols">
              <t>Utilize device lock to prevent unauthorized device access -
              <xref format="default" target="device_lock"></xref></t>

              <t>Combine refresh token requests with user-provided secret -
              <xref target="user_secret"></xref></t>

              <t>Where a user knows the device has been stolen, they can
              revoke the affected tokens - <xref
              target="refresh_revocation"></xref></t>
            </list> <spanx style="emph"></spanx></t>

          <t><spanx style="emph">Clone device: </spanx>All device data and
          applications are copied to another device. Applications are used
          as-is on the target device.</t>

          <t><spanx style="emph">Countermeasures:</spanx></t>

          <t><list style="symbols">
              <t>Combine refresh token request with device identification -
              <xref target="device_id"></xref></t>

              <t>Combine refresh token requests with user-provided secret -
              <xref target="user_secret"></xref></t>

              <t>Refresh Token Replacement - <xref
              target="refresh_replace"></xref></t>

              <t>Where a user knows the device has been cloned, they can use
              this countermeasure - Refresh Token Revocation - <xref
              target="refresh_revocation"></xref></t>
            </list> <spanx style="emph"></spanx></t>

          <t><spanx style="emph">Obtain refresh tokens from backup:</spanx> A
          refresh token could be obtained from the backup of a client
          application, or device.</t>

          <t><spanx style="emph">Countermeasures:</spanx></t>

          <t><list style="symbols">
              <t>tbd</t>
            </list></t>
        </section>

        <section title="Threat: Obtain Access Tokens">
          <t>Depending on the client type, there are different ways access
          tokens may be revealed to an attacker. Access tokens could be stolen
          from the device if the application stores them in a storage, which
          is accessible to other applications.</t>

          <t>Impact: Where the token is a bearer token and no additional
          mechanism is used to identify the client, the attacker can access
          all resources associated with the token and its scope.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Keep access tokens in transient memory and limit grants:
              <xref target="access_tokens"></xref></t>

              <t>Limited scope tokens - <xref target="limit_scope"></xref></t>

              <t>Keep access tokens in private memory or apply same protection
              means as for refresh tokens - <xref
              target="refresh_tokens"></xref></t>

              <t>Keep access token lifetime short - <xref
              target="short_exp_time"></xref></t>
            </list></t>
        </section>

        <section title="Threat: End-user credentials phished using compromised or embedded browser">
          <t>A malicious application could attempt to phish end-user passwords
          by misusing an embedded browser in the end-user authorization
          process, or by presenting its own user-interface instead of allowing
          trusted system browser to render the authorization user interface.
          By doing so, the usual visual trust mechanisms may be bypassed (e.g.
          TLS confirmation, web site mechanisms). By using an embedded or
          internal client application user interface, the client application
          has access to additional information it should not have access to
          (e.g. uid/password).</t>

          <t>Impact: If the client application or the communication is
          compromised, the user would not be aware and all information in the
          authorization exchange could be captured such as username and
          password.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Client developers and end-user can be educated to trust an
              external System-Browser only.</t>

              <t>Client applications could be validated prior publication in a
              application market.</t>

              <t>Client developers should not collect authentication
              information directly from users and should instead use redirects
              to and back from a trusted external system-browser.</t>
            </list></t>
        </section>
      </section>

      <section title="Authorization Endpoint">
        <t></t>

        <section title="Threat: Password phishing by counterfeit authorization server">
          <t>OAuth makes no attempt to verify the authenticity of the
          Authorization Server. A hostile party could take advantage of this
          by intercepting the Client's requests and returning misleading or
          otherwise incorrect responses. This could be achieved using DNS or
          ARP spoofing. Wide deployment of OAuth and similar protocols may
          cause Users to become inured to the practice of being redirected to
          websites where they are asked to enter their passwords. If Users are
          not careful to verify the authenticity of these websites before
          entering their credentials, it will be possible for attackers to
          exploit this practice to steal Users' passwords.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Authorization servers should consider such attacks when
              developing services based on OAuth, and should require
              transport-layer security for any requests where the authenticity
              of the authorization server or of request responses is an issue
              (see <xref target="server_authn"></xref>).</t>

              <t>Authorization servers should attempt to educate Users about
              the risks phishing attacks pose, and should provide mechanisms
              that make it easy for users to confirm the authenticity of their
              sites.</t>
            </list></t>
        </section>

        <section title="Threat: User unintentionally grants too much access scope">
          <t>When obtaining end user authorization, the end-user may not
          understand the scope of the access being granted and to whom or they
          may end up providing a client with access to resources which should
          not be permitted.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Explain the scope (resources and the permissions) the user is
              about to grant in a understandable way - <xref
              target="informed_decisions"></xref></t>

              <t>Narrow scope based on client-specific policy - When obtaining
              end user authorization and where the client requests scope, the
              authorization server may want to consider whether to honour that
              scope based on who the client is. That decision is between the
              client and authorization server and is outside the scope of this
              spec. The authorization server may also want to consider what
              scope to grant based on the profile used, e.g. providing lower
              scope where no client secret is provided from a native
              application. - <xref target="limit_scope"></xref></t>
            </list></t>
        </section>

        <section anchor="mal_client3"
                 title="Threat: Malicious client obtains existing authorization by fraud">
          <t>Authorization servers may wish to automatically process
          authorization requests from clients which have been previously
          authorized by the user. When the user is redirected to the
          authorization server's end-user authorization endpoint to grant
          access, the authorization server detects that the user has already
          granted access to that particular client. Instead of prompting the
          user for approval, the authorization server automatically redirects
          the user back to the client.</t>

          <t>A malicious client may exploit that feature and try to obtain
          such an authorization code instead of the legitimate client.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Authorization servers should not automatically process repeat
              authorizations where the client is not authenticated through a
              client secret or some other authentication mechanism such as
              signing with security certificates (see <xref
              target="strong_client_authn"></xref>) or validation of a
              pre-registered redirect URI (<xref target="val_redirect"></xref>
              )</t>

              <t>Authorization servers can mitigate the risks associated with
              automatic processing by limiting the scope of Access Tokens
              obtained through automated approvals - <xref
              target="limit_scope"></xref></t>
            </list></t>
        </section>

        <section title="Threat: Open redirector">
          <t>An attacker could use the end-user authorization endpoint and the
          redirect_uri parameter to abuse the authorization server as
          redirector.</t>

          <t>Impact?</t>

          <t>Countermeasure</t>

          <t><list style="symbols">
              <t>don't redirect to redirect_uri, if client identity or
              redirect_uri could not be verified</t>
            </list></t>
        </section>
      </section>

      <section title="Token endpoint">
        <t></t>

        <section title="Threat: Eavesdropping access tokens">
          <t>Attackers may attempts to eavesdrop access token on transit from
          the authorization server to the client.</t>

          <t>Impact: The attacker is able to access all resources with the
          permissions covered by the scope of the particular access token.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Authorization servers MUST ensure that these transmissions
              are protected using transport-layer mechanisms such as TLS or
              SSL (see <xref target="conf_requests"></xref>).</t>

              <t>If end-to-end confidentiality cannot be guaranteed, reducing
              scope (see <xref target="limit_scope"></xref>) and expiry time
              (<xref target="short_exp_time"></xref>) for access tokens can be
              used to reduce the damage in case of leaks.</t>
            </list></t>
        </section>

        <section title="Threat: Obtain access tokens from authorization server database">
          <t>This threat is applicable if the authorization server stores
          access tokens as handles in a database. An attacker may obtain
          access tokens from the authorization server's database by gaining
          access to the database or launching a SQL injection attack. Impact:
          disclosure of all access tokens</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>System security measures - <xref target="std_sys"></xref></t>

              <t>Store access token hashes only - <xref
              target="noclear"></xref></t>

              <t>Standard SQL injection Countermeasures - <xref
              target="std_sql"></xref></t>
            </list></t>
        </section>

        <section title="Threat: Obtain client credentials over non secure transport">
          <t>An attacker could attempt to eavesdrop the transmission of client
          credentials between client and server during the client
          authentication process or during OAuth token requests. Impact:
          Revelation of a client credential enabling the possibility for
          phishing or imitation of a client service.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Implement transport security through <xref format="title"
              target="conf_requests"></xref></t>

              <t>Alternative authentication means, which do not require to
              send plaintext credentials over the wire (Examples: Digest
              authentication)</t>
            </list></t>
        </section>

        <section title="Threat: Obtain client secret from authorization server database">
          <t>An attacker may obtain valid client_id/secret combinations from
          the authorization server's database by gaining access to the
          database or launching a SQL injection attack. Impact: disclosure of
          all client_id/secret combinations. This allows the attacker to act
          on behalf of legitimate clients.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Ensure proper handling of credentials as per <xref
              format="title" target="cred_storage_prot"></xref>.</t>
            </list></t>
        </section>

        <section title="Threat: Obtain client secret by online guessing">
          <t>An attacker may try to guess valid client_id/secret pairs.
          Impact: disclosure of single client_id/secret pair.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>High entropy of secrets - <xref
              target="high_entropy"></xref></t>

              <t>Lock accounts - <xref target="lock_accounts"></xref></t>
            </list></t>
        </section>

        <section title="Threat: DoS on dynamic client secret creation">
          <t>If an authorization servers includes a nontrivial amount of
          entropy in client secrets and if the authorization server
          automatically grants them, an attacker could exhaust the pool by
          repeatedly applying for them.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>The authorization server should consider some verification
              step for new clients. The authorization server should include a
              nontrivial amount of entropy in client secrets.</t>
            </list></t>
        </section>
      </section>

      <section title="Obtaining Authorization">
        <t>This section covers threats which are specific to certain flows
        utilized to obtain access tokens. Each flow is characterized by
        response types and/or grant types on the end-user authorization and
        tokens endpoint, respectively.</t>

        <section anchor="code_flow" title="Authorization Code">
          <section anchor="eavesdropping"
                   title="Threat: Eavesdropping or leaking authorization codes">
            <t>An attacker could try to eavesdrop transmission of the
            authorization code between authorization server and client.
            Furthermore, authorization codes are passed via the browser which
            may unintentionally leak those codes to untrusted web sites and
            attackers by different ways:</t>

            <t><list style="symbols">
                <t>Referrer headers: browsers frequently pass a
                &ldquo;referer&rdquo; header when a web page embeds content,
                or when a user travels from one web page to another web page.
                These referrer headers may be sent even when the origin site
                does not trust the destination site. The referee header is
                commonly logged for traffic analysis purposes.</t>

                <t>Request logs: web server request logs commonly include
                query parameters on requests.</t>

                <t>Open redirectors: web sites sometimes need to send users to
                another destination via a redirector. Open redirectors pose a
                particular risk to web-based delegation protocols because the
                redirector can leak verification codes to untrusted
                destination sites.</t>

                <t>Browser history: web browsers commonly record visited URLs
                in the browser history. Another user of the same web browser
                may be able to view URLs that were visited by previous
                users.</t>
              </list>Note: A description of a similar attacks on the SAML
            protocol can be found at
            http://www.oasis-open.org/committees/download.php/3405/oasis-sstc-saml-bindings-1.1.pdf
            (&sect;4.1.1.9.1),
            http://www.thomasgross.net/publications/papers/GroPfi2006-SAML2_Analysis_Janus.WSSS_06.pdf
            and
            http://www.oasis-open.org/committees/download.php/11191/sstc-gross-sec-analysis-response-01.pdf.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Authorization server as well as the client MUST ensure that
                these transmissions are protected using transport-layer
                mechanisms such as TLS or SSL (see <xref
                target="conf_requests"></xref>).</t>

                <t>The authorization server shall require the client to
                authenticate wherever possible, so the binding of the
                authorization code to a certain client can be validated in a
                reliable way (see <xref
                target="bind_code_client_id"></xref>).</t>

                <t>Limited duration of authorization codes - <xref
                target="short_exp_time"></xref></t>

                <t>The authorization server SHOULD enforce a one time usage
                restriction (see <xref target="one_time_usage"></xref>).</t>

                <t>If an Authorization Server observes multiple attempts to
                redeem a authorization code, the Authorization Server may want
                to revoke all tokens granted based on the authorization code
                (see <xref target="automatic_code_revocation"></xref>).</t>

                <t>In the absence of these countermeasures, reducing scope
                (<xref target="limit_scope"></xref>) and expiry time (<xref
                target="short_exp_time"></xref>) for access tokens can be used
                to reduce the damage in case of leaks.</t>

                <t>The client server may reload the target page of the
                redirect_uri in order to automatically cleanup the browser
                cache.</t>
              </list></t>
          </section>

          <section title="Threat: Obtain authorization codes from authorization server database">
            <t>This threat is applicable if the authorization server stores
            authorization codes as handles in a database. An attacker may
            obtain authorization codes from the authorization server's
            database by gaining access to the database or launching a SQL
            injection attack. Impact: disclosure of all authorization codes,
            most likely along with the respective redirect_uri and client_id
            values.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Credential storage protection can be employed - <xref
                target="cred_storage_prot"></xref></t>

                <t>System security measures - <xref
                target="std_sys"></xref></t>

                <t>Store access token hashes only - <xref
                target="noclear"></xref></t>

                <t>Standard SQL injection countermeasures - <xref
                target="std_sql"></xref></t>
              </list></t>
          </section>

          <section title="Threat: Online guessing of authorization codes">
            <t>An attacker may try to guess valid authorization code values
            and send it using the grant type &bdquo;code&ldquo; in order to
            obtain a valid access token. Impact: disclosure of single access
            token (+probably refresh token)</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>For handle-based designs: <xref
                target="random_entropy"></xref></t>

                <t>For assertion-based designs: <xref
                target="signed_tokens"></xref></t>

                <t>Authenticate the client, adds another value the attacker
                has to guess - <xref
                target="depl_specific_secretes"></xref></t>

                <t>Binding of authorization code to redirect_uri, adds another
                value the attacker has to guess - <xref
                target="bind_code_redirect"></xref></t>

                <t>Short expiration time - <xref
                target="short_exp_time"></xref></t>
              </list></t>
          </section>

          <section anchor="mal_client"
                   title="Threat: Malicious client obtains authorization">
            <t>A malicious client could counterfeit a valid client and obtain
            an access authorization that way. The malicious client could even
            utilize screen scraping techniques in order to simulate the user
            consent in the authorization flow.</t>

            <t>Assumption: It is not the task of the authorization server to
            protect the end-user's device from malicious software. This is the
            responsibility of the platform running on the particular device
            probably in cooperation with other components of the respective
            ecosystem (e.g. an application management infrastructure). The
            sole responsibility of the authorization server is to control
            access to the end-user's resources living in resource servers and
            to prevent unauthorized access to them. Based on this assumption,
            the following countermeasures are available to cope with the
            threat.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The authorization server should authentication the client,
                if possible (see <xref
                target="depl_specific_secretes"></xref>). Note: the
                authentication takes place after the end-user has authorized
                the access.</t>

                <t>The authorization server should validate the client's
                redirect_uri against the pre-registered redirect_uri, if one
                exists (see <xref target="val_redirect"></xref>). Note: The
                validation of the redirect_uri is the only technical mean to
                recognize a malicious client id in advance of the
                authorization process. Further note this does not work for
                native applications because in contrast to web applications
                this URI is not bound to a single communication endpoint. The
                valid client's redirect_uri (typically with custom scheme) can
                be used by a malicious on any device.</t>

                <t>After authenticating the end-user, the authorization server
                should ask him/her for consent. In this context, the user
                shall be explained the purpose, scope, and duration of the
                authorization. Moreover, the authorization server must view to
                the end-user the meta data it associates with the particular
                client. It is up to the user to validate the binding of this
                data to the particular application (e.g. Name) and to approve
                the authorization request. (see <xref
                target="validation_end_user"></xref>).</t>

                <t>The authorization server must not perform automatic
                re-authorizations for clients it is unable to reliably
                authenticate or validate (see <xref
                target="automatic_processing"></xref>).</t>

                <t>If the authorization server automatically authenticates the
                end-user, it may nevertheless require some user input in order
                to prevent screen scraping. Examples are CAPTCHAs or
                user-specific secret like PIN codes.</t>

                <t>The authorization server may also limit the scope of tokens
                it issues to clients it cannot reliably authenticate (see
                <xref target="limit_scope"></xref>).</t>
              </list></t>
          </section>

          <section title="Threat: Authorization code phishing">
            <t>A hostile party could impersonate the client site and get
            access to the authorization code. This could be achieved using DNS
            or ARP spoofing. This applies to clients, which are web
            applications, thus the redirect URI is not local to the host where
            the user's browser is running.</t>

            <t>Impact: This affects web applications and may lead to a
            disclosure of authorization codes and, potentially, the
            corresponding access and refresh tokens.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The authorization server MUST require the client to
                authenticate with a secret, so the binding of the
                authorization code to a certain client can be validated in a
                reliable way (see <xref
                target="bind_code_client_id"></xref>).</t>

                <t>The redirect_uri of the client SHOULD point to a HTTPS
                protected endpoint and the browser shall be utilized to
                authenticate this redirect_uri using server authentication
                (see <xref target="server_authn"></xref>).</t>
              </list></t>
          </section>

          <section title="Threat: User session impersonation">
            <t>A hostile party could impersonate the client site and
            impersonate the user's session on this client. This could be
            achieved using DNS or ARP spoofing. This applies to clients, which
            are web applications, thus the redirect URI is not local to the
            host where the user's browser is running.</t>

            <t>Impact: An attacker who intercepts the authorization code as it
            is sent by the browser to the callback endpoint can gain access to
            protected resources by submitting the authorization code to the
            client. The client will exchange the authorization code for an
            access token and use the access token to access protected
            resources for the benefit of the attacker, delivering protected
            resources to the attacker, or modifying protected resources as
            directed by the attacker. If OAuth is used by the client to
            delegate authentication to a social site (e.g. as in the
            implementation of the "Facebook Login" button), the attacker can
            use the intercepted authorization code to log in to the client as
            the user.</t>

            <t>Note: Authenticating the client during authorization code
            exchange will not help to detect such an attack as it is the
            legitimate client that obtains the tokens.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>In order to prevent an attacker from impersonating the
                end-users session, the redirect_uri of the client MUST point
                to a HTTPS protected endpoint and the browser shall be
                utilized to authenticate this redirect_uri using server
                authentication (see <xref target="server_authn"></xref>)</t>
              </list></t>
          </section>

          <section anchor="authz_code_leakage"
                   title="Threat: Authorization code leakage through counterfeit client">
            <t>The attack leverages the authorization code grant type in an
            attempt to get another user (victim) to log-in, authorize access
            to his/her resources, and subsequently obtain the authorization
            code and inject it into a client application using the attacker's
            account. The goal is to associate an access authorization for
            resources of the victim with the user account of the attacker on a
            client site.</t>

            <t>The attacker abuses an existing client application and combines
            it with his own counterfeit client web site. The attack depends on
            the victim expecting the client application to request access to a
            certain resource server. The victim, seeing only a normal request
            from an expected application, approves the request. The attacker
            then uses the victim's authorization to gain access to the
            information unknowingly authorized by the victim.</t>

            <t>The attacker conducts the following flow:</t>

            <t><list style="numbers">
                <t>The attacker accesses the client web site (or application)
                and initiates data access to a particular resource server. The
                client web site in turn initiates an authorization request to
                the resource server's authorization server. Instead of
                proceeding with the authorization process, the attacker
                modifies the authorization server end-user authorization URL
                as constructed by the client to include a redirect_uri
                parameter referring to a web site under his control
                (attacker's web site).</t>

                <t>The attacker tricks another user (the victim) to open that
                modified end-user authorization URI and to authorize access
                (e.g. an email link, or blog link). The way the attacker
                achieve that goal is out of scope.</t>

                <t>Having clicked the link, the victim is requested to
                authenticate and authorize the client site to have access.</t>

                <t>After completion of the authorization process, the
                authorization server redirects the user agent to the
                attacker's web site instead of the original client web
                site.</t>

                <t>The attacker obtains the authorization code from his web
                site by means out of scope of this document.</t>

                <t>He then constructs a redirect_uri to the target web site
                (or application) based on the original authorization request's
                redirect_uri and the newly obtained authorization code and
                directs his user agent to this URL. The authorization code is
                injected into the original client site (or application).</t>

                <t>The client site uses the authorization code to fetch a
                token from the authorization server and associates this token
                with the attacker's user account on this site.</t>

                <t>The attacker may now access the victims resources using the
                client site.</t>
              </list></t>

            <t>Impact: The attackers gains access to the victim's resources as
            associated with his account on the client site.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The attacker must use another redirect_uri for its
                authorization process than the target web site because it
                needs to intercept the flow. So if the authorization server
                associates the authorization code with the redirect_uri of a
                particular end-user authorization and validates this
                redirect_uri with the redirect_uri passed to the tokens
                endpoint, such an attack is detected (see <xref
                target="bind_code_redirect"></xref>).</t>

                <t>The authorization server may also enforce the usage and
                validation of pre-registered redirect URIs (see <xref
                target="val_redirect"></xref>). This will allow for an early
                recognition of session fixation attempts.</t>

                <t>For native applications, one could also consider to use
                deployment-specific client ids and secrets (see <xref
                target="depl_specific_secretes"></xref>, along with the
                binding of authorization code to client_id (see <xref
                target="bind_code_client_id"></xref>), to detect such an
                attack because the attacker does not have access the
                deployment-specific secret. Thus he will not be able to
                exchange the authorization code.</t>

                <t>The client may consider to use other flows, which are not
                vulnerable to this kind of attacks such as "Implicit Grant" or
                "Resource Owner Password Credentials" (see <xref
                target="implicite_flow"></xref> or <xref
                target="pwd_flow"></xref>).</t>
              </list></t>
          </section>

          <section title="Threat: CSRF attack against redirect-uri">
            <t>Cross-Site Request Forgery (CSRF) is a web-based attack whereby
            HTTP requests are transmitted from a user that the website trusts
            or has authenticated (e.g., via HTTP redirects or HTML forms).
            CSRF attacks on OAuth approvals can allow an attacker to obtain
            authorization to OAuth protected resources without the consent of
            the User.</t>

            <t>This attack works against the redirect-uri used in the
            authorization code flow. An attacker could authorize an
            authorization code to their own protected resources on an
            authorization server. He then aborts the redirect flow back to the
            client on his device and tricks the victim into executing the
            redirect back to the client. The client receives the redirect,
            fetches the token(s) from the authorization server and associates
            the victim's client session with the resources accessible using
            the token.</t>

            <t>Impact: The user accesses resources on behalf of the attacker.
            The effective impact depends on the type of resource accessed. For
            example, the user may upload private items to an attacker's
            resources. Or when using OAuth in 3rd party login scenarios, the
            user may associate his client account with the attacker's identity
            at the external identity provider. This way the attacker could
            easily access the victim's data at the client by logging in from
            another device with his credentials at the external identity
            provider.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The state parameter should be used to link the
                authorization request with the redirect-uri used deliver the
                access token. This will ensure the client is not tricked into
                completing any redirect callback unless it is linked to an
                authorization request the client initiated. The state
                parameter should be unguessable and the client should be
                capable of keeping the state parameter secret.</t>

                <t>Client developers and end-user can be educated not follow
                untrusted URLs.</t>
              </list></t>
          </section>

          <section title="Threat: Clickjacking attack against authorization">
            <t>With Clickjacking, a malicious site loads the target site in a
            transparent iframe overlaid on top of a set of dummy buttons which
            are carefully constructed to be placed directly under important
            buttons on the target site. When a user clicks a visible button,
            they are actually clicking a button (such as an "Authorize"
            button) on the hidden page.</t>

            <t>Impact: An attacker can steal a user's authentication
            credentials and access their resources</t>

            <t>Countermeasure</t>

            <t><list style="symbols">
                <t>Native applications SHOULD use external browsers instead of
                embedding browsers in an iFrame when requesting end-user
                authorization</t>

                <t>For newer browsers, avoidance of iFrames can be enforced
                server side by using the X-FRAME-OPTION header - <xref
                target="clickjacking_xframe"></xref></t>

                <t>For older browsers, javascript framebusting techniques can
                be used but may not be effective in all browsers.</t>
              </list></t>
          </section>

          <section title="Threat: DoS, Exhaustion of resources attacks">
            <t>If an authorization server includes a nontrivial amount of
            entropy in authorization codes or access tokens (limiting the
            number of possible codes/tokens) and automatically grants either
            without user intervention and has no limit on code or access
            tokens per user, an attacker could exhaust the pool by repeatedly
            directing user(s) browser to request code or access tokens. This
            is because more entropy means a larger number of tokens can be
            issued.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The authorization server should consider limiting the
                number of access tokens granted per user. The authorization
                server should include a nontrivial amount of entropy in
                authorization codes.</t>
              </list></t>
          </section>

          <section title="Threat: DoS using manufactured authorization codes">
            <t>An attacker who owns a botnet can locate the redirect URIs of
            clients that listen on HTTP, access them with random authorization
            codes, and cause a large number of HTTPS connections to be
            concentrated onto the authorization server. This can result in a
            DoS attack on the authorization server.</t>

            <t>This attack can still be effective even when CSRF defense/the
            'state' parameter are deployed on the client side. With such a
            defense, the attacker might need to incur an additional HTTP
            request to obtain a valid CSRF code/ state parameter. This
            apparently cuts down the effectiveness of the attack by a factor
            of 2. However, if the HTTPS/HTTP cost ratio is higher than 2 (the
            cost factor is estimated to be around 3.5x at
            http://www.semicomplete.com/blog/geekery/ssl-latency.html), the
            attacker still achieves a magnification of resource utilization at
            the expense of the authorization server.</t>

            <t>Impact: There are a few effects that the attacker can
            accomplish with this OAuth flow that they cannot easily achieve
            otherwise.</t>

            <t><list style="numbers">
                <t>Connection laundering: With the clients as the relay
                between the attacker and the authorization server, the
                authorization server learns little or no information about the
                identity of the attacker. Defenses such rate limiting on the
                offending attacker machines are less effective due to the
                difficulty to identify the attacking machines. Although an
                attacker could also launder its connections through an
                anonymizing systems such as Tor, the effectiveness of that
                approach depends on the capacity of the annoying system. On
                the other hand, a potentially large number of OAuth clients
                could be utilized for this attack.</t>

                <t>Asymmetric resource utilization: The attacker incurs the
                cost of an HTTP connection and causes an HTTPS connection to
                be made on the authorization server; and the attacker can
                co-ordinate the timing of such HTTPS connections across
                multiple clients relatively easily. Although the attacker
                could achieve something similar, say, by including an iframe
                pointing to the HTTPS URL of the authorization server in an
                HTTP web page and lure web users to visit that page, timing
                attacks using such a scheme may be more difficult as it seems
                nontrivial to synchronize a large number of users to
                simultaneously visit a particular site under the attacker's
                control.</t>
              </list>Countermeasures</t>

            <t><list style="symbols">
                <t>Though not a complete countermeasure by themselves, CSRF
                defense and the 'state' parameter created with secure random
                codes SHOULD be deployed on the client side. The client SHOULD
                forward the authorization code to the authorization server
                only after both the CSRF token and the 'state' parameter are
                validated.</t>

                <t>If the client authenticates the user, either through a
                single sign on protocol ( such as OpenID / Facebook Connect )
                or through local authentication, the client SHOULD suspend the
                access by a user account if the number of invalid
                authorization codes submitted by this user exceeds a certain
                threshold.</t>

                <t>The authorization server SHOULD send an error response to
                the client reporting an invalid authorization code and rate
                limit or disallow connections from clients whose number of
                invalid requests exceeds a threshold.</t>

                <t>The authorization server MAY in addition sign the
                authorization code using the public key from its SSL
                certificate, and require the client to validate the signature.
                To enhance interoperability between multiple clients and
                authorization servers, a standard procedure to create and
                validate the signature (including what attributes to sign) MAY
                be developed and agreed between the clients and the
                servers.</t>
              </list></t>
          </section>
        </section>

        <section anchor="implicite_flow" title="Implicit Grant">
          <t>he implicit grant flow, the access token is directly returned to
          the client as fragment part of the redirect_uri. It is assumed that
          the token is not send to the redirect_uri target since HTTP user
          agents do not send fragments server requests. Thus an attacker
          cannot eavesdrop the access token on this communication path and It
          cannot leak through HTTP referee headers.</t>

          <section title="Threat: Access token leak in transport/end-points">
            <t>This token might be eavesdropped by an attacker. The token is
            sent from server to client via a URI fragment of the redirect_uri.
            If the communication is not secured or the end-point is not
            secured, the token could be leaked by parsing the returned
            URI.</t>

            <t>Impact: the attacker would be able to assume the same rights
            granted by the token.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The authorization server must ensure confidentiality of the
                response from the authorization server to the client (see
                <xref target="conf_requests"></xref>).</t>
              </list></t>
          </section>

          <section title="Threat: Access token leak in browser history">
            <t>An attacker could obtain the token from the browsers history.
            Note this means the attacker needs access to the particular
            device.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Shorten token duration (see <xref
                target="short_exp_time"></xref>) and reduced scope of the
                token may reduce the impact of that attack (see <xref
                target="limit_scope"></xref>).</t>

                <t>Make these requests non-cachable</t>

                <t>Native applications can directly embed a browser widget and
                therewith gain full control of the cache. So the application
                can cleanup browser history after authorization process.</t>
              </list></t>
          </section>

          <section anchor="mal_client2"
                   title="Threat: Malicious client obtains authorization">
            <t>An malicious client could attempt to obtain a token by
            fraud.</t>

            <t>The same countermeasures as for <xref
            target="mal_client"></xref> are applicable, except client
            authentication.</t>
          </section>

          <section title="Threat: Manipulation of scripts">
            <t>A hostile party could act as the client web server and replace
            or modify the actual implementation of the client (script). This
            could be achieved using DNS or ARP spoofing. This applies to
            clients implemented within the Web Browser in a scripting
            language.</t>

            <t>Impact: The attacker could obtain user credential information
            and assume the full identity of the user.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The authorization server must authenticate the server from
                which scripts are obtained (see <xref
                target="server_authn"></xref>).</t>

                <t>The client must ensure that scripts obtained have not been
                altered in transport (see <xref
                target="conf_requests"></xref>).</t>

                <t>Introduce one time per-use secrets (e.g. client_secret)
                values that can only be used by scripts in a small time window
                once loaded from a server. The intention would be to reduce
                the effectiveness of copying client-side scripts for re-use in
                an attackers modified code. [[pending discussion]]</t>
              </list></t>
          </section>

          <section title="Threat: CSRF attack against redirect-uri">
            <t>Cross-Site Request Forgery (CSRF) is a web-based attack whereby
            HTTP requests are transmitted from a user that the website trusts
            or has authenticated (e.g., via HTTP redirects or HTML forms).
            CSRF attacks on OAuth approvals can allow an attacker to obtain
            authorization to OAuth protected resources without the consent of
            the User.</t>

            <t>This attack works against the redirect-uri used in the implicit
            grant flow. An attacker could acquire an access token to their own
            protected resources. He could then construct a Redirect-uri and
            embed their access token in that URI. If he can trick the user
            into following the Redirect-uri and the client does not have
            protection against this attack, the user may have the attacker's
            access token authorized within their client.</t>

            <t>Impact: The user accesses resources on behalf of the attacker.
            The effective impact depends on the type of resource accessed. For
            example, the user may upload private items to an attacker's
            resources. Or when using OAuth in 3rd party login scenarios, the
            user may associate his client account with the attacker's identity
            at the external identity provider. This way the attacker could
            easily access the victim's data at the client by logging in from
            another device with his credentials at the external identity
            provider.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>The state parameter should be used to link the
                authorization request with the redirect-uri used deliver the
                access token. This will ensure the client is not tricked into
                completing any redirect callback unless it is linked to an
                authorization request the client initiated. The state
                parameter should be unguessable and the client should be
                capable of keeping the state parameter secret.</t>

                <t>Client developers and end-user can be educated not follow
                untrusted URLs.</t>
              </list></t>
          </section>
        </section>

        <section anchor="pwd_flow" title="Resource Owner Password Credentials">
          <t>The &ldquo;Resource Owner Password Credentials&rdquo; grant type
          (see <xref target="I-D.ietf-oauth-v2"></xref>, Section 4.3), often
          used for legacy/migration reasons, allows a client to request an
          access token using an end-users user-id and password along with its
          own credential. This grant type has higher risk because it maintains
          the uid/password anti-pattern. Additionally, because the user does
          not have control over the authorization process, clients using this
          grant type are not limited by scope, but instead have potentially
          the same capabilities as the user themselves. As there is no
          authorization step, the ability to offer token revocation is
          bypassed.</t>

          <t>Impact: The resource server can only differentiate scope based on
          the access token being associated with a particular client. The
          client could also acquire long-living tokens and pass them up to a
          attacker web service for further abuse. The client, eavesdroppers,
          or end-points could eavesdrop user id and password.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Except for migration reasons, minimize use of this grant
              type</t>

              <t>The authorization server must validate the client id
              associated with the particular refresh token with every refresh
              request - <xref target="binding_refresh_client_id"></xref></t>

              <t>Authorization server MUST ensure that these transmissions are
              protected using transport-layer mechanisms such as TLS or SSL
              (see <xref target="conf_requests"></xref>).</t>
            </list></t>

          <section title="Threat: Accidental exposure of passwords at client site">
            <t>If the client does not provide enough protection, an attacker
            or disgruntled employee could retrieve the passwords for a
            user.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Use other flows, which do not rely on the client's
                cooperation for secure resource owner credential handling</t>

                <t>Use digest authentication instead of plaintext credential
                processing</t>

                <t>Obfuscation of passwords in logs</t>
              </list></t>
          </section>

          <section title="Threat: Client obtains scopes without end-user authorization">
            <t>All interaction with the resource owner is performed by the
            client. Thus it might, intentionally or unintentionally, happen
            that the client obtains a token with scope unknown for or
            unintended by the resource owner. For example, the resource owner
            might think the client needs and acquires read-only access to its
            media storage only but the client tries to acquire an access token
            with full access permissions.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Use other flows, which do not rely on the client's
                cooperation for resource owner interaction</t>

                <t>The authorization server may generally restrict the scope
                of access tokens (<xref target="limit_scope"></xref>) issued
                by this flow. If the particular client is trustworthy and can
                be authenticated in a reliable way, the authorization server
                could relax that restriction. Resource owners may prescribe
                (e.g. in their preferences) what the maximum permission for
                client using this flow shall be.</t>

                <t>The authorization server could notify the resource owner by
                an appropriate media, e.g. e-Mail, of the grant issued (see
                <xref target="informed"></xref>).</t>
              </list></t>
          </section>

          <section title="Threat: Client obtains refresh token through automatic authorization">
            <t>All interaction with the resource owner is performed by the
            client. Thus it might, intentionally or unintentionally, happen
            that the client obtains a long-term authorization represented by a
            refresh token even if the resource owner did not intend so.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Use other flows, which do not rely on the client's
                cooperation for resource owner interaction</t>

                <t>The authorization server may generally refuse to issue
                refresh tokens in this flow (see <xref
                target="restricted_refresh"></xref>). If the particular client
                is trustworthy and can be authenticated in a reliable way (see
                client authentication), the authorization server could relax
                that restriction. Resource owners may allow or deny (e.g. in
                their preferences) to issue refresh tokens using this flow as
                well.</t>

                <t>The authorization server could notify the resource owner by
                an appropriate media, e.g. e-Mail, of the refresh token issued
                (see <xref target="informed"></xref>).</t>
              </list></t>
          </section>

          <section title="Threat: Obtain user passwords on transport">
            <t>An attacker could attempt to eavesdrop the transmission of
            end-user credentials with the grant type &bdquo;password&ldquo;
            between client and server.</t>

            <t>Impact: disclosure of a single end-users password.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Confidentiality of Requests - <xref
                target="conf_requests"></xref></t>

                <t>alternative authentication means, which do not require to
                send plaintext credentials over the wire (Examples: Digest
                authentication)</t>
              </list></t>
          </section>

          <section title="Threat: Obtain user passwords from authorization server database">
            <t>An attacker may obtain valid username/password combinations
            from the authorization server's database by gaining access to the
            database or launching a SQL injection attack.</t>

            <t>Impact: disclosure of all username/password combinations. The
            impact may exceed the domain of the authorization server since
            many users tend to use the same credentials on different
            services.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Credential storage protection can be employed - <xref
                target="cred_storage_prot"></xref></t>
              </list></t>
          </section>

          <section title="Threat: Online guessing">
            <t>An attacker may try to guess valid username/password
            combinations using the grant type &bdquo;password&ldquo;.</t>

            <t>Impact: Revelation of a single username/password
            combination.</t>

            <t>Countermeasures:</t>

            <t><list style="symbols">
                <t>Password policy - <xref target="pwd_policy"></xref></t>

                <t>Lock accounts - <xref target="lock_accounts"></xref></t>

                <t>Tar pit - <xref target="tar_pit"></xref></t>

                <t>CAPTCHA - <xref target="captchas"></xref></t>

                <t>Abandon on grant type &bdquo;password&ldquo;</t>

                <t>Client authentication (see <xref
                target="client_aa"></xref>) will provide another
                authentication factor and thus hinder the attack.</t>
              </list></t>
          </section>
        </section>

        <section title="Client Credentials">
          <t>Client credentials (see <xref target="I-D.ietf-oauth-v2"></xref>,
          Section 3) consist of an identifier (not secret) combined with an
          additional means (such as a matching client secret) of
          authenticating a client.</t>

          <t>[[Threats seem to be covered elsewhere such as <xref
          target="eavesdropping"></xref>]]</t>
        </section>
      </section>

      <section title="Refreshing an Access Token">
        <section title="Threat: Eavesdropping refresh tokens from authorization server">
          <t>An attacker may eavesdrop refresh tokens when they are
          transmitted from the authorization server to the client.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Authorization servers MUST ensure that these transmissions
              are protected using transport-layer mechanisms such as TLS or
              SSL (see <xref target="conf_requests"></xref>).</t>

              <t>If end-to-end confidentiality cannot be guaranteed, reducing
              scope (see <xref target="limit_scope"></xref>) and expiry time
              (see <xref target="short_exp_time"></xref>) for issued access
              tokens can be used to reduce the damage in case of leaks.</t>
            </list></t>
        </section>

        <section title="Threat: Obtaining refresh token from authorization server database">
          <t>This threat is applicable if the authorization server stores
          refresh tokens as handles in a database. An attacker may obtain
          refresh tokens from the authorization server's database by gaining
          access to the database or launching a SQL injection attack.</t>

          <t>Impact: disclosure of all refresh tokens</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Credential storage protection - <xref
              target="cred_storage_prot"></xref></t>

              <t>Bind token to client id, if the attacker cannot obtain the
              required id and secret - <xref
              target="bind_token_client_id"></xref></t>
            </list></t>
        </section>

        <section title="Threat: Obtain refresh token by online guessing">
          <t>An attacker may try to guess valid refresh token values and send
          it using the grant type &bdquo;refresh_token&ldquo; in order to
          obtain a valid access token.</t>

          <t>Impact: exposure of single refresh token and derivable access
          tokens.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>For handle-based designs - <xref
              target="random_entropy"></xref></t>

              <t>For assertion-based designs - <xref
              target="signed_tokens"></xref></t>

              <t>Bind token to client id, because the attacker would guess the
              matching client id, too (see <xref
              target="bind_token_client_id"></xref>)</t>

              <t>Authenticate the client, adds another element the attacker
              has to guess (see <xref
              target="depl_specific_secretes"></xref>)</t>
            </list></t>
        </section>

        <section title="Threat: Obtain refresh token phishing by counterfeit authorization server">
          <t>An attacker could try to obtain valid refresh tokens by proxying
          requests to the authorization server. Given the assumption that the
          authorization server URL is well-known at development time or can at
          least be obtained from a well-known resource server, the attacker
          must utilize some kind of spoofing in order to succeed.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Server authentication (as described in <xref
              target="server_authn"></xref>)</t>
            </list></t>
        </section>
      </section>

      <section title="Accessing Protected Resources">
        <section title="Threat: Eavesdropping access tokens on transport">
          <t>An attacker could try to obtain a valid access token on transport
          between client and resource server. As access tokens are shared
          secrets between authorization and resource server, they MUST by
          treated with the same care as other credentials (e.g. end-user
          passwords).</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Access tokens sent as bearer tokens, SHOULD NOT be sent in
              the clear over an insecure channel. Instead transport protection
              means shall be utilized to prevent eavesdropping by an attacker
              (see <xref target="conf_requests"></xref>).</t>

              <t>A short lifetime reduces impact in case tokens are
              compromised (see <xref target="short_exp_time"></xref>).</t>

              <t>The access token can be bound to a client's identity and
              require the client to authenticate with the resource server (see
              <xref target="authn_requests"></xref>). Client authentication
              MUST be performed without exposing the required secret to the
              transport channel.</t>
            </list></t>
        </section>

        <section title="Threat: Replay authorized resource server requests">
          <t>An attacker could attempt to replay valid requests in order to
          obtain or to modify/destroy user data.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>The resource server should utilize transport security measure
              in order to prevent such attacks (see <xref
              target="conf_requests"></xref>). This would prevent the attacker
              from capturing valid requests.</t>

              <t>Alternatively, the resource server could employ signed
              requests (see <xref target="signed_requests"></xref>) along with
              nounces and timestamps in order to uniquely identify requests.
              The resource server MUST detect and refuse every replayed
              request.</t>
            </list></t>
        </section>

        <section title="Threat: Guessing access tokens">
          <t>Where the token is a handle, the attacker may use attempt to
          guess the access token values based on knowledge they have from
          other access tokens.</t>

          <t>Impact: Access to a single user's data.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Handle Tokens should have a reasonable entropy (see <xref
              target="random_entropy"></xref>) in order to make guessing a
              valid token value difficult.</t>

              <t>Assertion (or self-contained token ) tokens contents SHALL be
              protected by a digital signature (see <xref
              target="signed_tokens"></xref>).</t>

              <t>Security can be further strengthened by using a short access
              token duration (see <xref target="exp_time"></xref> and <xref
              target="short_exp_time"></xref>).</t>
            </list></t>
        </section>

        <section title="Threat: Access token phishing by counterfeit resource server">
          <t>An attacker may pretend to be a particular resource server and to
          accept tokens from a particular authorization server. If the client
          sends a valid access tokens to this counterfeit resource server, the
          server in turn may use that token to access other services on behalf
          of the resource owner.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Clients SHOULD not make authenticated requests with an access
              token to unfamiliar resource servers, regardless of the presence
              of a secure channel. If the resource server address is
              well-known to the client, it may authenticate the resource
              servers (see <xref target="server_authn"></xref>).</t>

              <t>Associate the endpoint address of the resource server the
              client talked to with the access token (e.g. in an audience
              field) and validate association at legitimate resource server.
              The endpoint address validation policy may be strict (exact
              match) or more relaxed (e.g. same host). This would require to
              tell the authorization server the resource server endpoint
              address in the authorization process.</t>

              <t>Associate an access token with a client and authenticate the
              client with resource server requests (typically via signature in
              order to not disclose secret to a potential attacker). This
              prevents the attack because the counterfeit server is assumed to
              miss the capabilities to correctly authenticate on behalf of the
              legitimate client to the resource server (<xref
              target="authn_requests"></xref>).</t>

              <t>Restrict the token scope (see <xref
              target="limit_scope"></xref>) and or limit the token to a
              certain resource server (<xref
              target="bind_token_rs"></xref>).</t>
            </list></t>
        </section>

        <section title="Threat: Abuse of token by legitimate resource server or client">
          <t>A legitimate resource server could attempt to use an access token
          to access another resource servers. Similarly, a client could try to
          use a token obtained for one server on another resource server.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Tokens should be restricted to particular resource servers
              (see <xref target="bind_token_rs"></xref>).</t>
            </list></t>
        </section>

        <section title="Threat: Leak of confidential data in HTTP-Proxies">
          <t>The HTTP Authorization scheme (OAuth HTTP Authorization Scheme)
          is optional. However, [RFC2616](Fielding, R., Gettys, J., Mogul, J.,
          Frystyk, H., Masinter, L., Leach, P., and T. Berners-Lee,
          &ldquo;Hypertext Transfer Protocol -- HTTP/1.1,&rdquo; .) relies on
          the Authorization and WWW-Authenticate headers to distinguish
          authenticated content so that it can be protected. Proxies and
          caches, in particular, may fail to adequately protect requests not
          using these headers. For example, private authenticated content may
          be stored in (and thus retrievable from) publicly-accessible
          caches.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Resource servers not using the HTTP Authorization scheme
              (OAuth HTTP Authorization Scheme - see <xref
              target="authz_header"></xref>) should take care to use other
              mechanisms, such as the Cache-Control header, to ensure that
              authenticated content is protected.</t>

              <t>Reducing scope (see <xref target="limit_scope"></xref>) and
              expiry time (<xref target="short_exp_time"></xref>) for access
              tokens can be used to reduce the damage in case of leaks.</t>
            </list></t>
        </section>

        <section title="Threat: Token leakage via logfiles and HTTP referrers">
          <t>If access tokens are sent via URI query parameters, such tokens
          may leak to log files and HTTP referrers.</t>

          <t>Countermeasures:</t>

          <t><list style="symbols">
              <t>Use authorization headers or POST parameters instead of URI
              request parameters (see <xref
              target="authz_header"></xref>).</t>

              <t>Set logging configuration appropriately</t>

              <t>Prevent unauthorized persons from access to system log files
              (see <xref target="std_sys"></xref>)</t>

              <t>HTTP referrers can be prevented by reloading the target page
              again without URI parameters</t>

              <t>Abuse of leaked access tokens can be prevented by enforcing
              authenticated requests (see <xref
              target="authn_requests"></xref>).</t>

              <t>The impact of token leakage may be reduced by limiting scope
              (see <xref target="limit_scope"></xref>) and duration (see <xref
              target="short_exp_time"></xref>) and enforcing one time token
              usage (see <xref target="one_time_usage"></xref>).</t>
            </list></t>
        </section>
      </section>
    </section>

    <section anchor="security_considerations" title="Security Considerations">
      <t>This section describes the countermeasures as recommended to mitigate
      the threats as described in Section 4.</t>

      <section title="General">
        <t>The general section covers consideratios that apply generally
        across all OAuth components (client, resource server, token server,
        and user-agents).</t>

        <section anchor="conf_requests" title="Confidentiality of Requests">
          <t>This is applicable to all requests sent from client to
          authorization server or resource server. While OAuth provides a
          mechanism for verifying the integrity of requests, it provides no
          guarantee of request confidentiality. Unless further precautions are
          taken, eavesdroppers will have full access to request content and
          may be able to mount interception or replay attacks through using
          content of request, e.g. secrets or tokens.</t>

          <t>Attacks can be mitigated by using transport-layer mechanisms such
          as TLS or SSL. VPN may considered as well.</t>

          <t>This is a countermeasure against the following threats:</t>

          <t><list style="symbols">
              <t>Replay of access tokens obtained on tokens endpoint or
              resource server's endpoint</t>

              <t>Replay of refresh tokens obtained on tokens endpoint</t>

              <t>Replay of authorization codes obtained on tokens endpoint
              (redirect?)</t>

              <t>Replay of user passwords and client secrets</t>
            </list></t>
        </section>

        <section anchor="server_authn" title="Server authentication">
          <t>HTTPS server authentication or similar means can be used to
          authenticate the identity of a server. The goal is to reliably bind
          the DNS name of the server to the public key presented by the server
          during connection establishment.</t>

          <t>The client MUST validate the binding of the server to its domain
          name. If the server fails to prove that binding, it is considered a
          man-in-the-middle attack. The security measure depends on the
          certification authorities the client trusts for that purpose.
          Clients should carefully select those trusted CAs and protect the
          storage for trusted CA certificates from modifications.</t>

          <t>This is a countermeasure against the following threats:</t>

          <t><list style="symbols">
              <t>Spoofing</t>

              <t>Proxying</t>

              <t>Phishing by counterfeit servers</t>
            </list></t>
        </section>

        <section anchor="informed"
                 title="Always keep the resource owner informed">
          <t>Transparency to the resource owner is a key element of the OAuth
          protocol. The user shall always be in control of the authorization
          processes and get the necessary information to meet informed
          decisions. Moreover, user involvement is a further security
          countermeasure. The user can probably recognize certain kinds of
          attacks better than the authorization server. Information can be
          presented/exchanged during the authorization process, after the
          authorization process, and every time the user wishes to get
          informed by using techniques such as:</t>

          <t><list style="symbols">
              <t>User consent forms</t>

              <t>Notification messages (e-Mail, SMS, &hellip;)</t>

              <t>Activity/Event logs</t>

              <t>User self-care applications or portals</t>
            </list></t>
        </section>

        <section title="Credentials">
          <t>This sections describes countermeasures used to protect all kinds
          of credentials from unauthorized access and abuse. Credentials are
          long term secrets, such as client secrets and user passwords as well
          as all kinds of tokens (refresh and access token) or authorization
          codes.</t>

          <section anchor="cred_storage_prot"
                   title="Credential storage protection">
            <section anchor="std_sys" title="Standard system security means">
              <t>A server system may be locked down so that no attacker may
              get access to sensible configuration files and databases.</t>
            </section>

            <section anchor="std_sql"
                     title="Standard SQL Injection Countermeasures">
              <t>If a client identifier or other authentication component is
              queried or compared against a SQL Database it may become
              possible for an injection attack to occur if parameters received
              are not validated before submission to the database.</t>

              <t><list style="symbols">
                  <t>Ensure that server code is using the minimum database
                  privileges possible to reduce the "surface" of possible
                  attacks.</t>

                  <t>Avoid dynamic SQL using concatenated input. If possible,
                  use static SQL.</t>

                  <t>When using dynamic SQL, parameterize queries using bind
                  arguments. Bind arguments eliminate possibility of SQL
                  injections.</t>

                  <t>Filter and sanitize the input. For example, if an
                  identifier has a known format, ensure that the supplied
                  value matches the identifier syntax rules.</t>
                </list></t>
            </section>

            <section anchor="noclear"
                     title="No cleartext storage of credentials">
              <t>The authorization server may consider to not store credential
              in clear text. Typical approaches are to store hashes instead.
              If the credential lacks a reasonable entropy level (because it
              is a user password) an additional salt will harden the storage
              to prevent offline dictionary attacks. Note: Some authentication
              protocols require the authorization server to have access to the
              secret in the clear. Those protocols cannot be implemented if
              the server only has access to hashes.</t>
            </section>

            <section title="Encryption of credentials">
              <t>For client applications, insecurely persisted client
              credentials are easy targets for attackers to obtain. Store
              client credentials using an encrypted persistence mechanism such
              as a keystore or database. Note that compiling client
              credentials directly into client code makes client applications
              vulnerable to scanning as well as difficult to administer should
              client credentials change over time.</t>
            </section>

            <section title="Use of asymmetric cryptography">
              <t>Usage of asymmetric cryptography will free the authorization
              server of the obligation to manage credentials. Nevertheless, it
              MUST ensure the integrity of the respective public keys.</t>
            </section>
          </section>

          <section anchor="online_secrets" title="Online attacks on secrets">
            <section anchor="pwd_policy" title="Password policy">
              <t>The authorization server may decide to enforce a complex user
              password policy in order to increase the user passwords'
              entropy. This will hinder online password attacks.</t>
            </section>

            <section anchor="high_entropy" title="High entropy of secrets">
              <t>When creating token handles or other secrets not intended for
              usage by human users, the authorization server MUST include a
              reasonable level of entropy in order to mitigate the risk of
              guessing attacks.</t>

              <t>The token value MUST be constructed from a cryptographically
              strong random or pseudo-random number sequence [RFC1750]
              generated by the Authorization Server. The probability of any
              two Authorization Code values being identical MUST be less than
              or equal to 2^(-128) and SHOULD be less than or equal to
              2^(-160).</t>
            </section>

            <section anchor="lock_accounts" title="Lock accounts">
              <t>Online attacks on passwords can be mitigated by locking the
              respective accounts after a certain number of failed
              attempts.</t>

              <t>Note: This measure can be abused to lock down legitimate
              service users.</t>
            </section>

            <section anchor="tar_pit" title="Tar pit">
              <t>The authorization server may react on failed attempts to
              authenticate by username/password by temporarily locking the
              respective account and delaying the response for a certain
              duration. This duration may increase with the number of failed
              attempts. The objective is to slow the attackers attempts on a
              certain username down.</t>

              <t>Note: this may require a more complex and stateful design of
              the authorization server.</t>
            </section>

            <section anchor="captchas" title="Usage of CAPTCHAs">
              <t>The idea is to prevent programs from automatically checking
              huge number of passwords by requiring human interaction.</t>

              <t>Note: this has a negative impact on user experience.</t>
            </section>
          </section>
        </section>

        <section title="Tokens (access, refresh, code)">
          <section anchor="limit_scope" title="Limit token scope">
            <t>The authorization server may decide to reduce or limit the
            scope associated with a token. Basis of this decision is out of
            scope, examples are:</t>

            <t><list style="symbols">
                <t>a client-specific policy, e.g. issue only less powerful
                tokens to unauthenticated clients,</t>

                <t>a service-specific policy, e.g. it a very sensible
                service,</t>

                <t>a resource-owner specific setting, or</t>

                <t>combinations of such policies and preferences.</t>
              </list>The authorization server may allow different scopes
            dependent on the grant type. For example, end-user authorization
            via direct interaction with the end-user (authorization code)
            might be considered more reliable than direct authorization via
            grant type username/password. This means will reduce the impact of
            the following threats:</t>

            <t><list style="symbols">
                <t>token leakage</t>

                <t>token issuance to malicious software</t>

                <t>unintended issuance of to powerful tokens with resource
                owner credentials flow</t>
              </list></t>
          </section>

          <section anchor="exp_time" title="Expiration time">
            <t>Tokens should generally expire after a reasonable duration.
            This complements and strengthens other security measures (such as
            signatures) and reduces the impact of all kinds of token
            leaks.</t>
          </section>

          <section anchor="short_exp_time" title="Short expiration time">
            <t>A short expiration time for tokens is a protection means
            against the following threats:</t>

            <t><list style="symbols">
                <t>replay</t>

                <t>reduce impact of token leak</t>

                <t>reduce likelihood of successful online guessing</t>
              </list>Note: Short token duration requires preciser clock
            synchronisation between authorization server and resource server.
            Furthermore, shorter duration may require more token refreshments
            (access token) or repeated end-user authorization processes
            (authorization code and refresh token).</t>
          </section>

          <section anchor="one_time_usage"
                   title="Limit number of usages/ One time usage">
            <t>The authorization server may restrict the number of request,
            which can be performed with a certain token. This mechanism can be
            used to mitigate the following threats:</t>

            <t><list style="symbols">
                <t>replay of tokens</t>

                <t>reduce likelihood of successful online guessing</t>
              </list>Additionally, If an Authorization Server observes
            multiple attempts to redeem a authorization code, the
            Authorization Server may want to revoke all tokens granted based
            on the authorization code.</t>
          </section>

          <section anchor="bind_token_rs"
                   title="Bind tokens to a particular resource server (Audience)">
            <t>Authorization servers in multi-service environments may
            consider to issue tokens with different content to different
            resource servers and to explicitly indicate in the token the
            target server a token is intended to be sent to (see Audience in
            SAML Assertions). This countermeasure can be used in the following
            situations:</t>

            <t><list style="symbols">
                <t>It reduce the impact of a successful replay attempt, since
                the token is applicable to a single resource server, only.</t>

                <t>It prevents abuse of a token by a rough resource server or
                client, since the token can only be used on that server. It is
                rejected by other servers.</t>

                <t>It reduce the impact of a leakage of a valid token to a
                counterfeit resource server.</t>
              </list></t>
          </section>

          <section anchor="endpoint_audience"
                   title="Use endpoint address as token audience">
            <t>This may be used to indicate to a resource server, which
            endpoint address has been used to obtain the token. This measure
            will allow to detect requests from a counterfeit resource server,
            since such token will contain the endpoint address of that
            server.</t>
          </section>

          <section anchor="audience_token_scope"
                   title="Audience and Token scopes">
            <t>Deployments may consider to use only tokens with explicitly
            defined scope, where every scope is associated with a particular
            resource server. This approach can be used to mitigate attacks,
            where a resource server or client uses a token for a different
            then the intended purpose.</t>
          </section>

          <section anchor="bind_token_client_id"
                   title="Bind token to client id">
            <t>An authorization server may bind a token to a certain client
            identity. This identity match must be validated for every request
            with that token. This means can be used, to</t>

            <t><list style="symbols">
                <t>detect token leakage and</t>

                <t>prevent token abuse.</t>
              </list>Note: Validating the client identity may require the
            target server to authenticate the client's identity. This
            authentication can be based on secrets managed independent of the
            token (e.g. pre-registered client id/secret on authorization
            server) or sent with the token itself (e.g. as part of the
            encrypted token content).</t>
          </section>

          <section anchor="signed_tokens" title="Signed tokens">
            <t>Self-contained tokens SHOULD be signed in order to detect any
            attempt to modify or produce faked tokens.</t>
          </section>

          <section anchor="enc_token" title="Encryption of token content">
            <t>Self-contained may be encrypted for privacy reasons or to
            protect system internal data.</t>
          </section>

          <section anchor="random_entropy"
                   title="Random token value with high entropy">
            <t>When creating token handles, the authorization server MUST
            include a reasonable level of entropy in order to mitigate the
            risk of guessing attacks. The token value MUST be constructed from
            a cryptographically strong random or pseudo-random number sequence
            [RFC1750] generated by the Authorization Server. The probability
            of any two token values being identical MUST be less than or equal
            to 2^(-128) and SHOULD be less than or equal to 2^(-160).</t>
          </section>

          <section title="Assertion formats">
            <t>For service providers intending to implement an assertion-based
            token design it is highly recommended to adopt a standard
            assertion format (such as SAML or JWT) that implements
            [draft-ietf-oauth-assertions].</t>
          </section>
        </section>

        <section anchor="access_tokens" title="Access tokens">
          <t><list style="symbols">
              <t>keep them in transient memory (accessible by the client
              application only)</t>

              <t>protect from exposure to 3rd parties (malicious
              application)</t>

              <t>limit number of access tokens granted to a user</t>
            </list></t>
        </section>
      </section>

      <section title="Authorization Server">
        <t>This section describes considerations related to the OAuth
        Authorization Server end-point.</t>

        <section title="Authorization Codes">
          <section anchor="automatic_code_revocation"
                   title="Automatic revocation of derived tokens if abuse is detected">
            <t>If an Authorization Server observes multiple attempts to redeem
            an authorization grant (e.g. such as an authorization code), the
            Authorization Server may want to revoke all tokens granted based
            on the authorization grant.</t>
          </section>
        </section>

        <section anchor="refresh_tokens" title="Refresh tokens">
          <section anchor="restricted_refresh"
                   title="Restricted issuance of refresh tokens">
            <t>The authorization server may decide based on an appropriate
            policy not to issue refresh tokens. Since refresh tokens are long
            term credentials, they may be subject theft. For example, if the
            authorization server does not trust a client to securely store
            such tokens, it may refuse to issue such a client a refresh
            token.</t>
          </section>

          <section anchor="binding_refresh_client_id"
                   title="Binding of refresh token to client_id">
            <t>The authorization server MUST bind every refresh token to the
            id of the client such a token was originally issued to and
            validate this binding for every request to refresh that token.
            This measure is a countermeasure against refresh token theft or
            leakage. </t>

            <t>Note: This binding MUST be protected from unauthorized
            modifications.</t>
          </section>

          <section anchor="refresh_replace" title="Refresh Token Replacement">
            <t>Refresh token replacement is intended to automatically detect
            and prevent attempts to use the same refresh token in parallel
            from different apps/devices. This happens if a token gets stolen
            from the client and is subsequently used by the attacker and the
            legitimate client. The basic idea is to change the refresh token
            value with every refresh request in order to detect attempts to
            obtain access tokens using old refresh tokens. Since the
            authorization server cannot determine whether the attacker or the
            legitimate client is trying to access, in case of such an access
            attempt the valid refresh token and the access authorization
            associated with it are both revoked.</t>

            <t>The OAuth specification supports this measure in that the
            tokens response allows the authorization server to return a new
            refresh token even for requests with grant type
            "refresh_token&ldquo;.</t>

            <t>Note: this measure may cause problems in clustered environments
            since usage of the currently valid refresh token must be ensured.
            In such an environment, other measures might be more
            appropriate.</t>
          </section>

          <section anchor="refresh_revocation"
                   title="Refresh Token Revocation">
            <t>The authorization server may allow clients or end-users to
            explicitly request the invalidation of refresh tokens.</t>

            <t>This is a countermeasure against:</t>

            <t><list style="symbols">
                <t>device theft,</t>

                <t>impersonation of resource owner, or</t>

                <t>suspected compromised client applications.</t>
              </list></t>
          </section>

          <section anchor="user_secret"
                   title="Combine refresh token requests with user-provided secret">
            <t>The exchange of a refresh token can be bound to the presence of
            a certain user-provided secret, such as a PIN, a password or a SIM
            card. This is a kind of multi-factor authentication on the tokens
            endpoint, since an attacker must possess both factors in order to
            be able to obtain an access token.</t>
          </section>

          <section anchor="device_id" title="Device identification">
            <t>The authorization server may require to bind authentication
            credentials to a device identifier or token assigned to that
            device. As the IMEI can be spoofed, that is not suitable, For
            mobile phones, a registration process can be used to assign a
            unique token to the device using an SMS message. That token or
            identifier can then be validated with when authenticating user
            credentials.</t>

            <t>This is a countermeasure against the following threats:</t>

            <t><list style="symbols">
                <t>phishing attacks</t>

                <t>...</t>
              </list></t>
          </section>

          <section anchor="clickjacking_xframe" title="X-FRAME-OPTION header">
            <t>For newer browsers, avoidance of iFrames can be enforced server
            side by using the X-FRAME-OPTION header. This header can have two
            values, deny and same origin, which will block any framing or
            framing by sites with a different origin, respectively.</t>

            <t>This is a countermeasure against the following threats:</t>

            <t><list style="symbols">
                <t>Clickjacking attacks</t>

                <t>...</t>
              </list></t>
          </section>
        </section>

        <section anchor="client_aa"
                 title="Public client authentication and authorization">
          <t>As described in Section 3 (Security Features), clients are
          identified, authenticated and authorized for several purposes, such
          as a</t>

          <t><list style="symbols">
              <t>Collate sub-sequent requests to the same client,</t>

              <t>Indicate the trustworthiness of a particular client to the
              end-user,</t>

              <t>Authorize access of clients to certain features on the
              authorization or resource server, and</t>

              <t>Log a client identity to log files for analysis or
              statics.</t>
            </list>Due to the different capabilities and characteristics of
          the different client types, there are different ways to support
          achieve objectives, which will be described in this section.
          Authorization server providers should be aware of the security
          policy and deployment of a particular clients and adapt its
          treatment accordingly. For example, one approach could be to treat
          all clients as less trustworthy and unsecure. On the other extreme,
          a service provider could activate every client installation by hand
          of an administrator and that way gain confidence in the identity of
          the software package and the security of the environment the client
          is installed in. And there are several approaches in between.</t>

          <section anchor="dont_issue"
                   title="Don't issue secrets to public clients or clients with inappropriate security policy">
            <t>Authorization servers should not issue secrets to "public"
            clients that cannot protect secrets. This prevents the server from
            overestimating the value of a successful authentication of the
            client.</t>

            <t>For example, it is of limited benefit to create a single client
            id and secret which is shared by all installations of a native
            application. Such a scenario requires that this secret must be
            transmitted from the developer via the respective distribution
            channel, e.g. an application market, to all installations of the
            application on end-user devices. A secret, burned into the source
            code of the application or a associated resource bundle, cannot be
            entirely protected from reverse engineering. Secondly, such
            secrets cannot be revoked since this would immediately put all
            installations out of work. Moreover, since the authorization
            server cannot really trust on the client's identity, it would be
            dangerous to indicate to end-users the trustworthiness of the
            client.</t>

            <t>There are other ways to achieve a reasonable security level, as
            described in the following sections.</t>
          </section>

          <section anchor="forced_user_consent"
                   title="Public clients without secret require user consent">
            <t>Authorization servers should not allow automatic authorization
            for public clients. The authorization may issue a client id, but
            SHOULD require that all authorizations are approved by the
            end-user. This is a countermeasure for clients without secret
            against the following threats:</t>

            <t><list style="symbols">
                <t>Impersonation of client applications using only
                client_id,</t>

                <t>Phishing attacks</t>

                <t>...</t>
              </list></t>
          </section>

          <section anchor="client_id_redirect"
                   title="Client_id only in combination with redirect_uri">
            <t>The authorization may issue a client_id and bind the client_id
            to a certain pre-configured redirect_uri. Any authorization
            request with another redirect_uri is refused automatically.
            Alternatively, the authorization server MUST not accept any
            dynamic redirect_uri for such a client_id and instead always
            redirect to the well-known pre-configured redirect_uri. This is a
            countermeasure for clients without secrets against the following
            threats:</t>

            <t><list style="symbols">
                <t>Cross-site scripting attacks</t>

                <t>Phishing attacks [...]</t>
              </list></t>
          </section>

          <section anchor="depl_specific_secretes"
                   title="Deployment-specific client secrets">
            <t>A authorization server may issue separate client identifiers
            and corresponding secrets to the different deployments of a
            client. The effect of such an approach would be to turn otherwise
            "public" clients back into "confidential" clients.</t>

            <t>For web applications, this could mean to create one client_id
            and client_secret per web site a software package is installed on.
            So the provider of that particular site could request client id
            and secret from the authorization server during setup of the web
            site. This would also allow to validate some of the properties of
            that web site, such as redirect_uri, address, and whatever proofs
            useful. The web site provider has to ensure the security of the
            client secret on the site.</t>

            <t>For native applications, things are more complicated because
            every installation of the application on any device is another
            deployment. Deployment specific secrets will require</t>

            <t><list style="numbers">
                <t>Either to obtain a client_id and client_secret during
                download process from the application market, or</t>

                <t>During installation on the device.</t>
              </list>Either approach will require an automated mechanism for
            issuing client ids and secrets, which is currently not defined by
            OAuth.</t>

            <t>The first approach would allow to achieve a level where the
            client is authenticated and identified, whereas the second option
            only allows to authenticate the client but not to validate
            properties of the client. But this would at least help to prevent
            several replay attacks. Moreover, deployment-specific client_id
            and secret allow to selectively revoke all refresh tokens of a
            specific deployment at once.</t>
          </section>

          <section anchor="val_redirect"
                   title="Validation of pre-registered redirect_uri">
            <t>An authorization server MUST require clients to register their
            redirect_uri or a pattern (TBD: make definition more precise)
            thereof. The way this registration is performed is out of scope of
            this document. Every actual redirect_uri sent with the respective
            client_id to the end-user authorization endpoint must comply with
            that pattern. Otherwise the authorization server must assume the
            inbound GET request has been sent by an attacker and refuse it.
            Note: the authorization server MUST NOT redirect the user agent
            back to the redirect_uri of such an authorization request.</t>

            <t><list style="symbols">
                <t>Authorization code leakage through counterfeit web site:
                allows to detect attack attempts already after first redirect
                to end-user authorization endpoint (<xref
                target="authz_code_leakage"></xref>).</t>

                <t>For clients with validated properties, this measure also
                helps to detect malicious applications early in the end-user
                authorization process. This reduces the need for a interactive
                validation by the user (<xref target="mal_client"></xref>,
                <xref target="mal_client2"></xref>).</t>
              </list>The underlying assumption of this measure is that an
            attacker must use another redirect_uri in order to get access to
            the authorization code. Deployments might consider the possibility
            of an attacker using spoofing attacks to a victims device to
            circumvent this security measure.</t>

            <t>Note: Pre-registering clients might not scale in some
            deployments (manual process) or require dynamic client
            registration (not specified yet). With the lack of dynamic client
            registration, it only works for clients bound to certain
            deployments at development/configuration time. As soon as dynamic
            resource server discovery gets involved, that's no longer
            feasible.</t>
          </section>

          <section anchor="client_secret_revocation"
                   title="Client secret revocation">
            <t>An authorization server may revoke a client's secret in order
            to prevent abuse of a revealed secret.</t>

            <t>Note: This measure will immediately invalidate any
            authorization code or refresh token issued to the respective
            client. This might be unintentionally for client identifiers and
            secrets used across multiple deployments of a particular native or
            web application.</t>

            <t>This a countermeasure against:</t>

            <t><list style="symbols">
                <t>...</t>

                <t>...</t>
              </list></t>
          </section>

          <section anchor="strong_client_authn"
                   title="Use strong client authentication (e.g. client_assertion / client_token)">
            <t>By using an alternative form of authentication such as client
            assertion [draft-ietf-oauth-assertions], the need to distribute
            client_secret is eliminated. This may require the use of a secure
            private key store or other supplemental authentication system as
            specified by the client assertion issuer in its authentication
            process.</t>
          </section>
        </section>

        <section title="End-user authorization">
          <t>This secion involves considerations for authorization flows
          involving the end-user.</t>

          <section anchor="automatic_processing"
                   title="Automatic processing of repeated authorizations requires client validation">
            <t>Authorization servers SHOULD NOT automatically process repeat
            authorizations where the client is not authenticated through a
            client secret or some other authentication mechanism such as
            signing with security certificates (5.7.2.7. Use strong client
            authentication (e.g. client_assertion / client_token)) or
            validation of a pre-registered redirect URI (5.7.2.5. Validation
            of pre-registered redirect_uri ).</t>
          </section>

          <section anchor="informed_decisions"
                   title="Informed decisions based on transparency">
            <t>The authorization server SHOULD clearly explain to the end-user
            what happens in the authorization process and what the
            consequences are. For example, the user shall understand what
            access he is about to grant to which client for what duration. It
            shall also be obvious to the user, whether the server is able to
            reliably certify certain client properties (web site address,
            security policy).</t>
          </section>

          <section anchor="validation_end_user"
                   title="Validation of client properties by end-user">
            <t>In the authorization process, the user is typically asked to
            approve a client's request for authorization. This is an important
            security mechanism by itself because the end-user can be involved
            in the validation of client properties, such as whether the client
            name known to the authorization server fits the name of the web
            site or the application the end-user is using. This measure is
            especially helpful in all situation where the authorization server
            is unable to authenticate the client. It is a countermeasure
            against:</t>

            <t><list style="symbols">
                <t>Malicious application</t>

                <t>A client application masquerading as another client</t>
              </list></t>
          </section>

          <section anchor="bind_code_client_id"
                   title="Binding of authorization code to client_id">
            <t>The authorization server MUST bind every authorization code to
            the id of the respective client which initiated the end-user
            authorization process. This measure is a countermeasure
            against:</t>

            <t><list style="symbols">
                <t>replay of authorization codes with different client
                credentials since an attacker cannot use another client_id to
                exchange an authorization code into a token</t>

                <t>Online guessing of authorization codes</t>
              </list>Note: This binding MUST be protected from unauthorized
            modifications.</t>
          </section>

          <section anchor="bind_code_redirect"
                   title="Binding of authorization code to redirect_uri">
            <t>The authorization server MUST bind every authorization code to
            the actual redirect_uri used as redirect target of the client in
            the end-user authorization process. This binding MUST be validated
            when the client attempts to exchange the respective authorization
            code for an access token. This measure is a countermeasure against
            authorization code leakage through counterfeit web sites since an
            attacker cannot use another redirect_uri to exchange an
            authorization code into a token.</t>
          </section>
        </section>
      </section>

      <section title="Client App Security">
        <t>This section deals with considerations for client applications.</t>

        <section anchor="cred_software"
                 title="Don't store credentials in code or resources bundled with software packages">
          <t>Because of the numbers of copies of client software, there is
          limited benefit to create a single client id and secret which is
          shared by all installations of an application. Such an application
          by itself would be considered a "public" client as it cannot be
          presumed to be able to keep client secrets. A secret, burned into
          the source code of the application or a associated resource bundle,
          cannot be entirely protected from reverse engineering. Secondly,
          such secrets cannot be revoked since this would immediately put all
          installations out of work. Moreover, since the authorization server
          cannot really trust on the client's identity, it would be dangerous
          to indicate to end-users the trustworthiness of the client.</t>
        </section>

        <section anchor="std_web"
                 title="Standard web server protection measures (for config files and databases)"></section>

        <section anchor="secure_storage"
                 title="Store secrets in a secure storage">
          <t>The are different way to store secrets of all kinds (tokens,
          client secrets) securely on a device or server.</t>

          <t>Most multi-user operation systems segregate the personal storage
          of the different system users. Moreover, most modern smartphone
          operating systems even support to store app-specific data in
          separate areas of the file systems and protect it from access by
          other applications. Additionally, applications can implements
          confidential data itself using a user-supplied secret, such as PIN
          or password.</t>

          <t>Another option is to swap refresh token storage to a trusted
          backend server. This mean in turn requires a resilient
          authentication mechanisms between client and backend server. Note:
          Applications must ensure that confidential data are kept
          confidential even after reading from secure storage, which typically
          means to keep this data in the local memory of the application.</t>
        </section>

        <section anchor="device_lock"
                 title="Utilize device lock to prevent unauthorized device access"></section>

        <section title="Platform security measures">
          <t><list style="symbols">
              <t>Validation process</t>

              <t>software package signatures</t>

              <t>Remote removal</t>
            </list></t>
        </section>
      </section>

      <section title="Resource Servers">
        <t>The following section details security considerations for resource
        servers.</t>

        <section anchor="authz_header" title="Authorization headers">
          <t>Authorization headers are recognized and specially treated by
          HTTP proxies and servers. Thus the usage of such headers for sending
          access tokens to resource servers reduces the likelihood of leakage
          or unintended storage of authenticated requests in general and
          especially Authorization headers.</t>
        </section>

        <section anchor="authn_requests" title="Authenticated requests">
          <t>An authorization server may bind tokens to a certain client
          identity and encourage resource servers to validate that binding.
          This will require the resource server to authenticate the originator
          of a request as the legitimate owner of a particular token. There
          are a couple of options to implement this countermeasure:</t>

          <t><list style="symbols">
              <t>The authorization server may associate the distinguished name
              of the client with the token (either internally or in the
              payload of an self-contained token). The client then uses client
              certificate-based HTTP authentication on the resource server's
              endpoint to authenticate its identity and the resource server
              validates the name with the name referenced by the token.</t>

              <t>same as before, but the client uses his private key to sign
              the request to the resource server (public key is either
              contained in the token or sent along with the request)</t>

              <t>Alternatively, the authorization server may issue a
              token-bound secret, which the client uses to sign the request.
              The resource server obtains the secret either directly from the
              authorization server or it is contained in an encrypted section
              of the token. That way the resource server does not "know" the
              client but is able to validate whether the authorization server
              issued the token to that client</t>
            </list>This mechanisms is a countermeasure against abuse of tokens
          by counterfeit resource servers.</t>
        </section>

        <section anchor="signed_requests" title="Signed requests">
          <t>A resource server may decide to accept signed requests only,
          either to replace transport level security measures or to complement
          such measures. Every signed request must be uniquely identifiable
          and must not be processed twice by the resource server. This
          countermeasure helps to mitigate:</t>

          <t><list style="symbols">
              <t>modifications of the message and</t>

              <t>replay attempts</t>
            </list></t>
        </section>
      </section>
    </section>

    <section anchor="IANA" title="IANA Considerations">
      <t>This document makes no request of IANA.</t>

      <t>Note to RFC Editor: this section may be removed on publication as an
      RFC.</t>
    </section>

    <section anchor="Acknowledgements" title="Acknowledgements">
      <t>We would like to thank Hui-Lan Lu, Francisco Corella, Eric Pflam,
      Shane B Weeden, Skylar Woodward and James H. Manger for their comments
      and contributions.</t>
    </section>
  </middle>

  <back>
    <references title="Normative References">
      <?rfc include="reference.RFC.2119"?>

      <?rfc include="reference.I-D.ietf-oauth-v2.xml"?>
    </references>

    <references title="Informative References">
      <?rfc include="reference.I-D.lodderstedt-oauth-revocation.xml"?>

      <?rfc include="reference.I-D.ietf-oauth-v2-bearer.xml"?>

      <?rfc include="reference.I-D.ietf-oauth-v2-http-mac.xml"?>

      <reference anchor="portable-contacts">
        <front>
          <title>Portable Contacts 1.0 Draft C</title>

          <author fullname="Joseph Smarr" initials="J." surname="Smarr">
            <organization></organization>
          </author>

          <date day="5" month="August" year="2008" />
        </front>
      </reference>
    </references>

    <section title="Document History">
      <t>[[ to be removed by RFC editor before publication as an RFC ]]</t>

      <t>draft-lodderstedt-oauth-security-01</t>

      <t><list style="symbols">
          <t>section 4.4.1.2 - changed "resource server" to "client" in
          countermeasures description.</t>

          <t>section 4.4.1.6 - changed "client shall authenticate the server"
          to "The browser shall be utilized to authenticate the redirect_uri
          of the client"</t>

          <t>section 5 - general review and alignment with public/confidential
          client terms</t>

          <t>all sections - general clean-up and typo corrections</t>
        </list></t>

      <t>draft-ietf-oauth-v2-threatmodel-00</t>

      <t><list style="symbols">
          <t>section 3.4 - added the purposes for using authorization
          codes.</t>

          <t>extended section 4.4.1.1</t>

          <t>merged 4.4.1.5 into 4.4.1.2</t>

          <t>corrected some typos</t>

          <t>reformulated "session fixation", renamed respective sections into
          "authorization code disclosure through counterfeit client"</t>

          <t>added new section "User session impersonation"</t>

          <t>worked out or reworked sections 2.3.3, 4.4.2.4, 4.4.4, 5.1.4.1.2,
          5.1.4.1.4, 5.2.3.5</t>

          <t>added new threat "DoS using manufactured authorization codes" as
          proposed by Eric Pflam</t>

          <t>added XSRF and clickjacking (incl. state parameter
          explanation)</t>

          <t>changed sub-section order in section 4.4.1</t>

          <t>incorporated feedback from Skylar Woodward (client secrets) and
          Shane B Weeden (refresh tokens as client instance secret)</t>

          <t>aligned client section with core draft's client type
          definition</t>

          <t>converted I-D into WG document</t>
        </list></t>
    </section>
  </back>
</rfc>