Skip to content
AI.info

The Pulse

EU asks AI deployers to test a draft rights-impact questionnaire under the AI Act

The European Commission is inviting public comments on a draft template for assessing the fundamental-rights impact of certain high-risk AI systems. It is also seeking deployers to test the questionnaire on their own systems.

EU asks AI deployers to test a draft rights-impact questionnaire under the AI Act

The European Commission is seeking public feedback until 15 January 2027 on a draft questionnaire for assessing how certain high-risk AI systems may affect fundamental rights. It also wants organisations that will be subject to the assessment requirement to test the draft on their own systems.

The exercise concerns a requirement under the EU AI Act to assess potential impacts before deploying certain high-risk systems. The assessment obligation applies from 2 December 2027.

A questionnaire for deployers

The AI Office developed the draft with input from the EU Agency for Fundamental Rights. The questionnaire follows the structure of Article 27 of the AI Act and sets out questions intended to guide deployers through the assessment.

A draft explanatory notice accompanies the template. It addresses which deployers need to carry out an assessment, when they need to do so and how their obligation relates to those of the high-risk system’s provider.

Once the template is finalised, deployers will need to use it to prepare their assessments and notify the results to the market surveillance authority. The Commission also plans to make it available as an automated tool on the AI Act Single Information Platform after adoption.

What the assessment covers

Article 27 covers specified deployers of high-risk systems, including bodies governed by public law and private entities providing public services, as well as deployers of certain systems identified in Annex III of the Act. The article excludes systems intended for use in the area listed under point 2 of that annex.

That area covers AI systems intended for use as safety components in the management and operation of critical digital infrastructure, road traffic, or the supply of water, gas, heating or electricity.

Under Article 27, assessments must describe the processes in which the system will be used, as well as the intended period and frequency of use. They must also identify the people and groups likely to be affected and the specific risks of harm to them, taking account of information supplied by the provider.

The assessment must describe human oversight measures and what will be done if the identified risks materialise, including internal governance arrangements and complaint mechanisms. The requirement applies to the first use of the high-risk system. Deployers must update the information in the template if they consider that an element of the assessment has changed or is no longer current during use.

Testing the draft

Deployers interested in testing the questionnaire are asked to express their interest to the AI Office by 4 December 2026. The call is directed at deployers that will have to carry out an assessment under Article 27.

Depending on the number of expressions of interest, the office may select cases for testing. It intends to ensure a balanced representation of areas and cases of use, types and sizes of deployers, and member states.

Participants will apply the template to one of their own systems. The AI Office will provide an introductory session, answer questions during testing and hold a closing session to discuss the results. Participation is voluntary and does not affect obligations under the AI Act.

Sources

  1. European Commission
  2. AI Act Service Desk
  3. AI Act Service Desk

Related stories

Explore

More articles