AI predicting sickness absence: almost always a legal no-go
Adopted 2026-06-21 · ≈ 2 min read · 
edited by Dirk Baaijen
AI predicting which workers will fall ill or be absent touches health data — special-category personal data with a strict prohibition regime (Art. 9 GDPR). The risk of discriminating against sick or disabled workers is high. In most cases such prediction is not lawful.
Short answer: AI predicting who will soon be ill or absent is one of the most sensitive HR applications — and almost always a legal no-go. It touches health data, special-category personal data with a strict prohibition regime (Art. 9 GDPR), and the risk of discriminating against sick or disabled workers is high. For most employers the honest answer is: don't.
Why it is so sensitive
Absence is linked to health, and health data falls under Article 9 GDPR: processing is in principle prohibited, save strict exceptions that almost never apply in an ordinary employment relationship. An employer may not even ask about the nature of an illness; a model predicting absence effectively infers health information you may not even hold.
The discrimination risk
An absence prediction implicitly penalises anyone with a chronic illness or disability. Use such a score in scheduling, evaluation or contract renewal and that is discrimination on grounds of health or disability — separate from the GDPR breach. It also touches the ban on exploiting vulnerability in the AI Act.
The AI Act on top
If the outcome steers decisions about workers (deployment, evaluation, renewal), the system is high-risk — on top of the GDPR problem. And a "wellbeing" or "vitality" tool that in fact infers emotions or health can fall under the emotion ban.
What is allowed
Anonymised, aggregated analysis at organisation level — for example to improve workload or safety risks — can be legitimate, provided it is not traceable to individuals and not used for decisions about people. The line is hard: as soon as it becomes individual and steering, it tips into unlawful.
What to do
- Do not predict individual absence — the GDPR basis is almost always missing.
- Do not use absence scores in scheduling, evaluation or contract decisions.
- Limit yourself to aggregated, non-traceable insights aimed at causes, not persons.
- Involve the occupational physician and the works council; absence is their domain, with its own safeguards.
The temptation is understandable — absence costs money. But an algorithm predicting who falls ill collects information the law protects and penalises people for their health. That is exactly what the rules aim to prevent.
Sources
- https://eur-lex.europa.eu/eli/reg/2016/679/oj
Regulation (EU) 2016/679 (GDPR): Art. 9 prohibits processing of health data except under strict exceptions. - https://eur-lex.europa.eu/eli/reg/2024/1689/oj
Regulation (EU) 2024/1689 (AI Act): high-risk once the outcome steers decisions about workers; Art. 5 where it exploits vulnerability.
Read next
AI background checks and social-media screening of candidates
AI that screens candidates via social media or background checks quickly clashes with the GDPR: proportionality, special-category data and transparency. Add reliability risks (false matches) and discrimination through irrelevant private information.
GDPR Article 88 and employee data: what does it mean for AI at work?
GDPR Article 88 lets Member States set their own rules for processing in the employment context. The Netherlands has no specific Art. 88 law, so the general GDPR plus the Dutch Implementation Act apply. With the weak basis of consent, purpose limitation and the role of works councils.
AI productivity scoring and bossware: what may an employer do?
Continuous AI monitoring and performance scoring ("bossware") collides with GDPR proportionality, purpose limitation and transparency, with good employership and with the works council's consent right. Covert or all-encompassing measurement is not allowed.