AI Policy

Principles guiding AI use and data handling during the Incubator

This page complements the Participation & Data Sharing Consent Agreement signed by every team and their PI.

Responsible and ethical use

AI tools used during the Program, including the Ersilia Model Hub and LLMs, support scientific judgment rather than replace it. Any AI-generated output must be reviewed and validated by the researcher before use in a project or publication, following the human oversight and accountability principles of the AU Continental AI Strategy aligned with the AU's Agenda 2063.

Vocabulary and transparency

Open-source is used strictly per the Open Source Initiative's definition. Content substantially generated or assisted by AI (text, code, figures) should be clearly identified as such, consistent with the transparency principles of the EU AI Act.

Ersilia's AI/ML tools and any other collaborator AI/ML tools using during the AI Incubator remain governed by their respective open-source licenses.

Regarding the Ersilia Model Hub, any model developed by teams during the Incubator will also be released as open-source.