chatPsych

About

chatPsych is a Python based, open-source interface for better human–AI research. It leverages widespread AI models and aims to improve data collection, experimental manipulations, and novel insights into AI–human interactions.

This tool will be introduced in a 2025 behavioural methods research paper:
“chatPsych: An Artificial Intelligence Interface for Experimental Human-AI Interaction.”

Developed by Oliver C. Lack at the Australian Institute for Machine Learning and the School of Psychology, University of Adelaide, chatPsych integrates a large selection of AI models through APIs into a GUI interface for online and local experimentation. It affords researchers an open-source and accessible method for experimental designs that use prompt engineering, model hyperparameters, multi-modalities interfaces, model comparison, and content moderation.

chatPsych facilitates collection of a plethora of interaction data especially relevant to various human-AI interaction research. It may be easily scaled and customised for prospective behavioural research projects.

This interface aims to make interaction with real-world AI systems more accessible. Integration with Qualtrics, Prolific, MTurk or other platforms for online sampling is easy.

The incorporation of such systems into generalisable experiments is imperative for human-AI research!

For usage, questions or collaborations, please cite and contact:

Oliver Lack
oliver.lack@adelaide.edu.au | oliver@oliverlack.com
oliverlack.com
Researcher Profile

Playground

Try out the chatPsych interface using a default AI agent (Prolific: user; Password: default).

Additionally, try a gamified 20-questions "Wordgame" interaction context (Prolific: user; Password: wordgame).

PLAYGROUND INTEGRATION STILL IN PROGRESS

Downloads

The chatPsych framework, including documentation and examples, is available on GitHub. Researchers can adapt or extend the code, incorporate new AI providers, or integrate specialized experimental features. It is free and open-source, encouraging community-driven improvements.

Follow the repository’s instructions for server deployment, local testing, recommended environment setup, SSL certification, and database maintenance.

Example adaptions like Wordie-AI also demonstrate how to run multi-condition experiments.

License

MIT License

Copyright (c) 2025 Oliver Lack

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Versions / Adaptions

Wordie-AI demonstrates an experiment that assigns participants to AI agent files with different temperatures, models, and system messages. It implements a 'Wordgame', similar to an open-ended version of 20-questions.

Developers

Lead: Oliver C. Lack

Institutional Support: Australian Institute for Machine Learning & School of Psychology, University of Adelaide

We welcome collaboration and pull requests. Driving innovation through open-source practices ensures the platform grows with new AI features and research needs.

For usage, questions or collaborations, please cite and contact:

Oliver Lack
oliver.lack@adelaide.edu.au | oliver@oliverlack.com
oliverlack.com
Researcher Profile

Blog

Our blog will feature usage guides, updates on newly integrated AI models, tutorials for lab-based or online expansions, and posts highlighting researchers’ findings.

We encourage the community to contact us about adaptions, updates, and research successes.