Recruitment & AI

Is AI in recruitment legal in Europe?

Yes, AI in recruitment is legal in Europe — provided you comply with GDPR, the AI Act, and national labour law. It is not a grey zone: it is a demanding but clear framework that many European employers already apply successfully.

This article answers the most common HR leader objection: "Are we allowed to do this?" Spoiler: yes, if you do it right.

Symbol of legality and compliance for AI in European recruitment.
Legality depends not on AI itself, but on how you deploy it.

Short answer

AI in hiring is permitted in Europe. It is governed by GDPR (personal data), the AI Act (high-risk systems), and labour law (non-discrimination, transparency). Prohibited uses include abusive profiling, fully automated decisions without human recourse, and certain invasive biometric or emotion claims.

The European legal framework in three pillars

1. GDPR (since 2018)

All candidate data is personal data. Inform, justify legal basis, limit retention, guarantee rights.

2. AI Act (since 2024, full effect 2026)

Algorithmic hiring systems are high-risk. See our AI Act for recruiters in 2026 article.

3. Labour law and non-discrimination

Algorithms that reproduce historical bias can engage employer liability — even without discriminatory intent.

Diagram of three legal pillars governing AI in European recruitment.
GDPR + AI Act + labour law: complementary frameworks, one rigour requirement.

What is permitted

  • AI voice or async interviews with prior candidate information;
  • Scoring on explicit job criteria defined by recruiters;
  • Automatic answer summaries to speed human review;
  • AI-assisted pre-screening with final human validation;
  • Soft skills analysis from behavioural answers (not opaque emotion engines);
  • Controlled retention of recordings with limited duration and security.

What is prohibited or strongly discouraged

  • Automatic rejection without human contestation;
  • Emotion or biometric scoring without solid justification;
  • Generalised profiling beyond role assessment;
  • Data collection without legal basis or candidate information;
  • Using sensitive data to filter candidates;
  • Subliminal manipulation (banned by the AI Act).

Five conditions to stay lawful

  1. Inform candidates before the AI interview.
  2. Document evaluation criteria and scoring logic.
  3. Supervise: a human validates every pass/reject decision.
  4. Secure data (encryption, EU hosting, Article 28 contract).
  5. Audit outcomes regularly for discriminatory effects.

Go further

For operational GDPR guidance, see AI-assisted recruitment: how to stay GDPR compliant and our GDPR and AI in recruitment overview. HiLucy is built for this legal framework — explore the platform.

Want to move from reading to action? See how Hi Lucy automates your voice AI interviews and your approach to interviews powered by artificial intelligence.

← Articles