For hiring managers, EU AI Act compliance means: candidates must be told when AI is assessing them; every AI-generated score must be reviewable, explainable, and overridable by a human; emotion recognition AI is prohibited; and a full audit trail of scoring decisions must be maintained. The requirements apply to all EU employers using AI in any part of the hiring process, with enforcement expected from December 2027.
What hiring managers must do
The EU AI Act creates specific obligations at the point where AI interacts with candidates. Before any AI assessment begins, candidates must be clearly informed: that AI is being used, what it is evaluating, and how the results will influence the hiring decision. This is not a legal boilerplate requirement — the disclosure must be specific and meaningful.
After assessment, candidates have the right to request human review of any AI-generated score. Hiring managers must be prepared to review and, where warranted, override AI outputs — and this review must be genuine, not a rubber-stamping exercise. The right to human review is a substantive obligation.
What is prohibited
Emotion recognition AI in employment contexts is explicitly prohibited. This covers any system that analyses facial expressions, vocal tone, micro-expressions, or physiological signals to infer emotional state, personality, or suitability for a role. If your current video interview platform analyses anything other than the content of candidates' responses, it is likely non-compliant.
Automated rejection — where AI eliminates a candidate without any human review — is also contrary to the Act's human oversight requirements. AI can rank and shortlist; it cannot make final hire or no-hire decisions without a human decision-maker in the loop.
Audit trail obligations
Every AI-assisted hiring decision must be documented: what was assessed, how it was scored, what the output was, and that a human reviewed the AI output before any decision was made. These records must be retained for a period sufficient to allow audit or candidate challenge — typically aligned with GDPR retention periods of three to six months for unsuccessful candidates.
For hiring managers, this means working with platforms that generate and retain these records automatically. Manually trying to reconstruct an audit trail after the fact is neither practical nor sufficient.
How Palantrix supports hiring manager compliance
Palantrix handles the compliance infrastructure automatically: candidate disclosure before the interview begins, transcript-only AI analysis, human-reviewable and overridable scores, and a complete audit trail retained for every hiring decision. Hiring managers using Palantrix operate within a compliant framework by default — they do not need to build or manage the compliance layer separately.
See how AI Video Interviews work →Frequently Asked Questions
When does the EU AI Act apply to hiring?
The high-risk AI provisions covering employment and recruitment are expected to be enforceable from December 2, 2027, following a provisional political agreement in May 2026 to extend the original August 2026 deadline. Formal adoption is pending. Organisations should continue compliance preparations regardless — the requirements themselves have not changed.
Does the EU AI Act apply to Irish employers?
Yes. The EU AI Act applies to all organisations operating in the EU, including Ireland. Irish employers using AI in hiring are subject to the Act's requirements, regardless of where the AI platform provider is headquartered.
What if we use AI for CV screening but not for interviews?
CV screening AI that ranks or filters candidates for employment decisions is also classified as high-risk under the EU AI Act. The same transparency, human oversight, and audit trail requirements apply. If any AI system is used to sort, rank, or evaluate candidates at any stage of the process, it is in scope.