Regulation
EU AI Act for assessment AI: a checklist for universities
AI used for admissions and for evaluating learning outcomes is listed as high-risk under the EU AI Act. The obligations now apply from 2 December 2027—time to plan, not to wait.
Where assessment AI sits
The AI Act lists, in Annex III point 3, AI systems used to decide access or admission to education (3(a)) and to evaluate learning outcomes (3(b)). Whether a specific tool is high-risk depends on its intended use and the legal conditions in the Act, so institutions should ask each vendor how they classify their system.
The timeline
High-risk rules were originally due in August 2026. The AI Omnibus, which entered into force on 27 July 2026, moved the date for Annex III systems—explicitly including education, per the European Parliament—to 2 December 2027.
Checklist for choosing assessment AI
- Risk management: is there a documented, ongoing risk process?
- Data governance: what data trained and tested the system, and is your students’ work kept out of training?
- Technical documentation and record-keeping: are assessments logged so results can be reviewed later?
- Transparency: can staff and students be told clearly how the tool is used?
- Human oversight: can a person review and override every result?
- Accuracy, robustness and cybersecurity: has it been tested on your own cohort against expert markers?
- Fairness: are results reported by language background and cohort?
This guide is general information, not legal advice. CERENO is designed towards these high-risk requirements; this is a design objective, not a claim of completed certification or conformity.