Article 5 of the EU AI Act: Prohibited AI Practices Explained
Article 5 sets the EU AI Act's hard red lines: AI uses considered an unacceptable risk to fundamental rights are banned outright, not merely regulated. These prohibitions have applied since 2 February 2025 and carry the Act's highest penalties.
What Article 5 requires
Article 5 prohibits, among others: harmful manipulation and exploitation of vulnerabilities; social scoring by public or private actors; untargeted scraping of facial images to build recognition databases; emotion recognition in the workplace and education (save narrow medical/safety uses); certain biometric categorisation by sensitive attributes; and real-time remote biometric identification in public spaces for law enforcement, except in narrowly authorised situations.
Who it binds
The prohibitions apply to everyone in scope — providers and deployers alike. There is no size exemption. Breaching Article 5 attracts fines of up to €35 million or 7% of global annual turnover, whichever is higher.
Key points
- In force since 2 February 2025 — the earliest deadline in the Act.
- Highest penalty tier: up to €35M or 7% of global turnover.
- Emotion recognition is banned at work and in education specifically.
- Real-time remote biometric ID by police is banned except in narrow, authorised cases.
FAQ
Is facial recognition banned under Article 5?
Untargeted scraping of facial images to build databases is banned, and real-time remote biometric identification in public spaces for law enforcement is banned except in narrow authorised cases. Other biometric identification is high-risk, not prohibited.
When did Article 5 take effect?
The Article 5 prohibitions have applied since 2 February 2025.
Related articles
Run the free 2-minute checker — deterministic, no sign-up.