The European Commission has published a draft form for the fundamental rights impact assessment (FRIA) and opened a public consultation on it, according to Digital Watch Observatory. The form was drawn up by the Commission's AI Office with input from the EU Agency for Fundamental Rights, and the deadline for submitting comments is 15 January 2027.
The assessment obligation, as stated, applies to certain categories of users of high-risk artificial intelligence systems: public bodies and private entities providing public services, users of systems for creditworthiness assessment and credit scoring, as well as those using artificial intelligence for risk assessment and pricing in life and health insurance. Not all users of high-risk systems are covered.
The draft questionnaire guides users through seven sections: general information about the organisation and the system, the context of use, affected groups, impacts and risks to fundamental rights with safeguards, human oversight, measures in the event of risks materialising, and a concluding section. A practical guide accompanies the draft, along with draft explanatory notes specifying who is covered, when the assessment is carried out and how it relates to the data protection impact assessment.
Under Article 27, covered users must carry out the assessment before the first use of the system and update it if relevant factors change, and submit the results to the competent national market surveillance authority. For certain public authorities and EU institutions, registration of a summary of findings in the public high-risk system database is also envisaged. The Commission invites users to apply by 4 December 2026 for voluntary testing of the questionnaire, and the obligations under Article 27 are due to apply from 2 December 2027.
What this means for businesses
For companies in the lending and insurance sectors, as well as for public service providers, this is an early signal that compliance with Article 27 will require a documented internal procedure rather than a one-off assessment. Taking part in testing the form may reduce the risk of having to adjust processes later, but does not guarantee the final shape of the form. Organisations should plan for updating assessments and notifying supervisory authorities as a regular part of managing high-risk artificial intelligence systems.
This summary was written with AI based on the cited source and checked by automated rules. For exact details, read the original article.