EU AI Act: high-risk HR obligations land August 2 — as Brussels debates a deferral
From August 2, AI used for recruitment, candidate selection, performance evaluation, task allocation, monitoring, and promotion/termination decisions is regulated as high-risk in the EU — with deployer duties (human oversight, worker notice, logging) and fines up to €15M or 3% of global turnover. Meanwhile the proposed Digital/AI Omnibus would defer some high-risk obligations — proposed, not passed.
Sources: DLA Piper · Ogletree · Crowell & Moring