9 November 2026

AI in recruitment: the deadline moved, the bans already apply

Software that filters CVs, ranks candidates or assesses interviews falls under the strictest category of the European AI Act. Those rules were due to apply on 2 August 2026. With the Digital Omnibus, Regulation 2026/1744, in force since 27 July, the EU moved that date to 2 December 2027. Part of the Act has applied since February 2025, though.

What already applies

Since 2 February 2025 a number of uses are banned. For recruitment the key one is the ban on emotion recognition in the workplace, and so in the job interview. A system that infers from a video whether a candidate is nervous, enthusiastic or sincere is not allowed.

The AI literacy duty started on the same date. The Digital Omnibus softened it. Employers must now take measures that support the AI knowledge of their staff. They do not have to guarantee a specific level for each employee.

What comes in December 2027

AI for recruitment and selection is listed in Annex III of the Act, the list of high-risk uses. From 2 December 2027 an employer using such a system must use it according to the instructions, keep human oversight, monitor how it works and keep its logs. The heaviest duties lie with the software provider.

A system you buy now still has to comply in December 2027.

What you can do

  • List which software in your recruitment selects, ranks or assesses anything, including a function inside a larger package.
  • Check whether any of those systems infers emotions or personality from video or voice, and switch that function off.
  • Ask vendors how they will comply before December 2027, and record the answer.
  • Record the measures for the AI knowledge of recruiters and hiring managers, even though they are now less strict.

Source: Regulation (EU) 2026/1744 (Digital Omnibus on AI), Official Journal 24 July 2026.