relational-ai

Code for the CHI 2026 paper: Scaffolded Vulnerability: Chatbot-Mediated Reciprocal Self-Disclosure and Need-Supportive Interaction in Couples

Overview

Echo is a Telegram-based chatbot designed to support meaningful self-disclosure between partners or close friends.
It guides users through structured conversation phases, encourages reflection on each other’s responses, and helps facilitate supportive interaction.

This repository contains the research prototype used in our study.

Features

• Structured conversation phases for guided self-disclosure
• LLM-based responses for warm and adaptive conversational support
• Partner reflection prompts that encourage supportive engagement
• Conversation summarization during or after the interaction
• Telegram group-chat integration for two-person conversations

Repository Structure

relational-ai/
├── main.py           # Main Telegram bot logic
├── prompts.py        # Prompt templates and phase instructions
├── requirements.txt  # Project dependencies
├── README.md
└── .env              # Local configuration file (not committed)

Requirements

• Python 3.10+
• A Telegram bot token from @BotFather
• An OpenAI API key

Installation

1. Clone the repository

git clone https://github.com/AMAAI-Lab/relational-ai.git
cd relational-ai

2. Create and activate a virtual environment

python -m venv .venv
source .venv/bin/activate

On Windows:

.venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

Configuration

Create a .env file in the project root:

OPENAI_API_KEY=your_openai_api_key
TELEGRAM_TOKEN=your_telegram_bot_token
API_MODEL=gpt-4o
TELEGRAM_BOT_NAME=@your_bot_username

Running the Bot

python main.py

Usage

1. Add the bot to a Telegram group chat with two users.
2. Start the bot with:
   • /start to begin
   • /continue to move to the next phase
   • /pause to pause the interaction
3. Follow the prompts in the group chat.

Notes

• This repository provides a research prototype and is not intended as a production deployment.
• The bot assumes a Telegram-based, two-user conversation setting.
• Local message logs may be stored during execution; make sure these are excluded from version control.

Citation

@inproceedings{10.1145/3772318.3791370,
author = {Jiang, Zhuoqun and Yeo, ShunYi and Herremans, Dorien and Tangi Perrault, Simon},
title = {Scaffolded Vulnerability: Chatbot-Mediated Reciprocal Self-Disclosure and Need-Supportive Interaction in Couples},
year = {2026},
isbn = {9798400722783},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3772318.3791370},
doi = {10.1145/3772318.3791370},
abstract = {While reciprocal self-disclosure drives intimacy, digital tools seldom scaffold autonomy, competence, and relatedness—the motivational underpinnings defined by Self-Determination Theory (SDT) that enable deep exchange. We introduce a chatbot employing dual-layer scaffolding to satisfy these needs: first providing enabling affordances (instrumental support) for vulnerability, then mediating affordances (relational support) for responsiveness. In a randomized study (N = 72; 36 couples) comparing Partner Support (PS: both layers), Direct Support (DS: enabling only), and Basic Prompt (BP: questions only), results reveal a critical distinction. While enabling affordances (PS, DS) were sufficient to deepen disclosure, only mediating affordances (PS) reliably elicited partner-provided need support and increased perceived closeness. Furthermore, controlled motivation decreased across conditions, and scaffolding buffered vitality, which remained stagnant in BP. We contribute empirical evidence that SDT-guided mediation fosters connection, offering a practical framework for designing AI-mediated conversations that support, rather than replace, human intimacy.},
booktitle = {Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems},
articleno = {1296},
numpages = {39},
keywords = {Human-human interaction, Self-disclosure, Conversational agent, Social computing, Relational technology},
location = {
},
series = {CHI '26}
}