The EU AI Act is the first comprehensive law regulating artificial intelligence, organised by the level of risk a system poses. Scheria uses AI to help researchers describe and match their work — it proposes, it does not decide. This page explains the Act's structure and where Scheria sits within it.
What the EU AI Act does
The Act takes a risk-based approach: obligations scale with the risk an AI system poses to people. The higher the potential for harm, the stricter the requirements; systems that pose little or no risk carry few or none. Every in-scope AI system is sorted into one of four categories, from prohibited at the top down to minimal at the bottom.
The risk tiers
Unacceptable risk: banned outright. High risk: strict obligations such as risk management, data governance, and human oversight. Limited risk: transparency duties under Article 50 — people must be told when they are dealing with an AI system. Minimal risk: no specific obligations beyond existing law. Most everyday AI tools sit in the limited- or minimal-risk tiers.
Where Scheria sits
Scheria's assistant is a limited-risk system, and its duty is transparency under Article 50: users are told when they are interacting with AI and when content is AI-assisted. The assistant proposes drafts and matches; humans decide. Responsibility splits at the platform boundary. Scheria is responsible for its own AI features; researchers remain responsible for the scientific content and the downstream use of their work. Scheria's intake includes a risk-classification step that flags work which might fall under stricter tiers, so it can be handled appropriately — as currently classified under Article 50.