Recruitment chatbots: transparency duty and high-risk once they select
Adopted 2026-06-21 · ≈ 2 min read · 
edited by Dirk Baaijen
A recruitment chatbot must always disclose that it is AI (Art. 50). Once it pre-selects, scores or rejects candidates, it is also a high-risk system (Annex III), with human oversight and the GDPR on top. A "handy assistant" thus quickly becomes a decision tool.
Short answer: A recruitment chatbot has two layers of regulation. The transparency duty always applies: the candidate must know they are talking to AI (Article 50). And as soon as the chatbot does more than inform — pre-select, score answers, pass or reject candidates — it is a high-risk system under Annex III, with all the duties that entails. The GDPR runs in parallel.
Layer 1: always transparency
Article 50 requires an AI system communicating directly with people to disclose that it is AI. A chatbot that poses as human — or leaves it ambiguous — does not comply. The disclosure must be clear and up front, not buried in the terms.
Layer 2: high-risk once it selects
Many recruitment chatbots do not stop at answering questions. They ask qualifying questions, score the answers and decide whether a candidate proceeds. At that point the chatbot performs selection — and falls under Annex III, point 4. Then:
- Human oversight (Art. 14): a recruiter assesses and can reverse the outcome; no automatic rejection based on a chat.
- Relevant, representative operation and monitoring for bias.
- Information to the candidate that a high-risk system is used.
The GDPR and accessibility
A rejection coming solely from the chatbot engages Article 22 GDPR: basis, transparency and a right to human intervention. Mind accessibility too: a chatbot that copes poorly with non-native speakers or people with a disability can exclude unintentionally — an extra discrimination risk.
The trap: "just an assistant"
The appeal of a recruitment chatbot is speed, and that is exactly where the risk sits: the more it pre-filters, the more it decides. Treat any chatbot that influences who proceeds as a selection tool, not a gadget.
What to do
- Have the chatbot disclose it is AI — clearly and up front.
- Determine whether it selects: if so, treat it as high-risk.
- Build a human decision point before every rejection.
- Test for bias and accessibility and record the result.
A chatbot that only informs is lightly regulated. A chatbot that selects is not. The difference is not the technology but what you let it decide.
Sources
- https://eur-lex.europa.eu/eli/reg/2024/1689/oj
Regulation (EU) 2024/1689 (AI Act): Art. 50 (chatbot discloses it is AI) and Annex III point 4 (selection of candidates = high-risk). - https://eur-lex.europa.eu/eli/reg/2016/679/oj
Regulation (EU) 2016/679 (GDPR): lawful basis and Art. 22 where the chatbot pre-selects or rejects.
Read next
AI in recruitment and HR: what every employer needs to know
AI in recruitment, selection and workforce management falls under Annex III of the AI Act and counts as high-risk — for every employer, regardless of sector or size. Emotion recognition in the workplace is banned, AI literacy already applies, and the GDPR runs in parallel for automated decisions.
AI in onboarding and internal mobility: where is the line?
Talent marketplaces, skills matching and career paths with AI seem neutral, but they reach the high-risk line as soon as they steer promotion or progression decisions (Annex III, point 4). Then the AI Act, GDPR, transparency and equal opportunity apply internally too.
AI in housing allocation: access to essential services
AI that decides who gets access to housing strikes at the core of the high-risk regime. Annex III of the AI Act covers access to essential private and public services; on top of that, the GDPR prohibits discrimination and sets demands on automated decisions.