What the EU AI Act means for hiring tools
From 2026, assessment tools used in hiring fall under 'high-risk' AI rules. Here is what employers and vendors must now be able to show.
Dr. Anna Lindqvist2 June 2026 · 8 minThe EU AI Act classifies AI systems used in recruitment and selection as 'high-risk' — and the obligations that come with that classification begin to bite in 2026. Combined with New York City's Local Law 144 and long-standing EEOC guidance, the direction is unambiguous: organisations must be able to evidence that their people decisions are fair, valid and defensible.
In practice that means documented validity, monitored adverse impact, transparency to candidates, and human oversight. Most employers cannot produce this today — not because their tools are bad, but because the evidence was never assembled.
This is precisely the gap the GTI Standard is built to close: a single, auditable benchmark that tells a board, a candidate, and a regulator that a decision can stand up.
The Global Talent Institute is the independent authority on fair, valid and defensible people decisions. Read the Standard or see how to get recognised.