diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index e240ae1cc3..49dfb6f65d 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -106,3 +106,9 @@ repos:
             types: [file, python]
             exclude: ^kedro/template/
             entry: isort
+          - id: bandit
+            name: "Bandit security check"
+            language: system
+            types: [file, python]
+            exclude: ^kedro/template/|^tests/
+            entry: bandit -ll
diff --git a/kedro/extras/datasets/pandas/gbq_dataset.py b/kedro/extras/datasets/pandas/gbq_dataset.py
index 3284a587b1..b8e75dee13 100644
--- a/kedro/extras/datasets/pandas/gbq_dataset.py
+++ b/kedro/extras/datasets/pandas/gbq_dataset.py
@@ -142,7 +142,7 @@ def _describe(self) -> Dict[str, Any]:
         )
 
     def _load(self) -> pd.DataFrame:
-        sql = "select * from {}.{}".format(self._dataset, self._table_name)
+        sql = "select * from {}.{}".format(self._dataset, self._table_name)  # nosec
         return pd.read_gbq(
             sql,
             project_id=self._project_id,
diff --git a/kedro/extras/datasets/pickle/pickle_dataset.py b/kedro/extras/datasets/pickle/pickle_dataset.py
index d3a1843a8c..7d40114e89 100644
--- a/kedro/extras/datasets/pickle/pickle_dataset.py
+++ b/kedro/extras/datasets/pickle/pickle_dataset.py
@@ -154,7 +154,7 @@ def _load(self) -> Any:
         load_path = get_filepath_str(self._get_load_path(), self._protocol)
 
         with self._fs.open(load_path, mode="rb") as fs_file:
-            return pickle.loads(fs_file.read(), **self._load_args)
+            return pickle.loads(fs_file.read(), **self._load_args)  # nosec
 
     def _save(self, data: Any) -> None:
         save_path = get_filepath_str(self._get_save_path(), self._protocol)
diff --git a/kedro/extras/datasets/spark/spark_hive_dataset.py b/kedro/extras/datasets/spark/spark_hive_dataset.py
index be22a3dcc5..0fe43ea58e 100644
--- a/kedro/extras/datasets/spark/spark_hive_dataset.py
+++ b/kedro/extras/datasets/spark/spark_hive_dataset.py
@@ -75,15 +75,16 @@ def __init__(
 
     def __enter__(self):
         self._data.createOrReplaceTempView("tmp")
+
         self._spark_session.sql(
-            "create table {stage_database_name}.{stage_table_name} as select * from tmp".format(
-                stage_database_name=self._stage_database_name,
+            "create table {stage_db_name}.{stage_table_name} as select * from tmp".format(  # nosec
+                stage_db_name=self._stage_database_name,
                 stage_table_name=self._stage_table_name,
             )
         ).take(1)
         self.staged_data = self._spark_session.sql(
-            "select * from {stage_database_name}.{stage_table_name}".format(
-                stage_database_name=self._stage_database_name,
+            "select * from {stage_db_name}.{stage_table_name}".format(  # nosec
+                stage_db_name=self._stage_database_name,
                 stage_table_name=self._stage_table_name,
             )
         )
@@ -91,8 +92,8 @@ def __enter__(self):
 
     def __exit__(self, exc_type, exc_val, exc_tb):
         self._spark_session.sql(
-            "drop table {stage_database_name}.{stage_table_name}".format(
-                stage_database_name=self._stage_database_name,
+            "drop table {stage_db_name}.{stage_table_name}".format(
+                stage_db_name=self._stage_database_name,
                 stage_table_name=self._stage_table_name,
             )
         )
@@ -208,12 +209,12 @@ def _get_spark() -> SparkSession:
     def _create_empty_hive_table(self, data):
         data.createOrReplaceTempView("tmp")
         self._get_spark().sql(
-            "create table {database}.{table} select * from tmp limit 1".format(
+            "create table {database}.{table} select * from tmp limit 1".format(  # nosec
                 table=self._table, database=self._database
             )
         )
         self._get_spark().sql(
-            "truncate table {database}.{table}".format(
+            "truncate table {database}.{table}".format(  # nosec
                 database=self._database, table=self._table
             )
         )
@@ -226,7 +227,7 @@ def _load(self) -> DataFrame:
                 )
             )
         return self._get_spark().sql(
-            "select * from {database}.{table}".format(
+            "select * from {database}.{table}".format(  # nosec
                 database=self._database, table=self._table
             )
         )
@@ -246,7 +247,7 @@ def _save(self, data: DataFrame) -> None:
     def _insert_save(self, data: DataFrame) -> None:
         data.createOrReplaceTempView("tmp")
         self._get_spark().sql(
-            "insert into {database}.{table} select {columns} from tmp".format(
+            "insert into {database}.{table} select {columns} from tmp".format(  # nosec
                 database=self._database,
                 table=self._table,
                 columns=", ".join(self._table_columns),
@@ -282,7 +283,7 @@ def _overwrite_save(self, data: DataFrame) -> None:
         self._get_spark().sql(
             "truncate table {database}.{table}".format(
                 database=self._database, table=self._table
-            )
+            )  # nosec
         )
         self._insert_save(data)
 
diff --git a/test_requirements.txt b/test_requirements.txt
index 96f037abce..ed6bf77cd3 100644
--- a/test_requirements.txt
+++ b/test_requirements.txt
@@ -2,6 +2,7 @@
 azure-storage-blob>=1.1.0, <2.0
 azure-storage-file>=1.1.0, <2.0
 azure-storage-queue>=1.1.0, <2.0
+bandit>=1.6.2, <2.0
 behave==1.2.6
 biopython>=1.73, <2.0
 black==v19.10.b0; python_version >= '3.6'
diff --git a/tools/ipython/ipython_loader.py b/tools/ipython/ipython_loader.py
index 6ae65dbe04..8d07710068 100644
--- a/tools/ipython/ipython_loader.py
+++ b/tools/ipython/ipython_loader.py
@@ -108,7 +108,7 @@ def run_startup_scripts(startup_dir: pathlib.Path):
                 compiled = compile(
                     script.read_text(encoding="utf-8"), str(script), "exec"
                 )
-                exec(compiled, globals())  # pylint: disable=exec-used
+                exec(compiled, globals())  # pylint: disable=exec-used # nosec
             except Exception as err:  # pylint: disable=broad-except
                 logging.error(
                     "Startup script `%s` failed:\n%s: %s",