Précis imagines what happens after the EU AI Act requires every member state to run a national AI regulatory sandbox. Right now, a tech founder testing an AI system in Luxembourg has to coordinate across data protection, standards, and sector regulators with no shared view of their case. Every new application starts from zero, with no visibility into how a comparable system was classified before.

Précis is a shared precedent layer. Founders see the closest prior sandbox decisions, and the reasoning behind them, before they submit anything. Regulators see the same precedent the moment a new case lands, instead of assessing risk cold. A joint timeline keeps the founder and every institution on one thread rather than three separate inboxes.

The tool is built to never decide anything itself but only ever surfaces precedent for a human to weigh. That constraint is the actual argument of the project: that responsible AI in a regulatory context starts with being explicit about what the system is and isn't allowed to do. This is a self-initiated concept project.