Overview
Article 5 of the EU AI Act prohibits certain AI practices that are considered to pose an unacceptable risk to fundamental rights, safety, and democratic values. These prohibitions took effect on February 2, 2025 — the first enforcement milestone of the AI Act. Violations carry the highest penalties: up to €35 million or 7% of global annual turnover.
Banned AI Practices
- Social Scoring: AI systems used by public authorities to evaluate or classify individuals based on social behavior, leading to detrimental treatment
- Subliminal or Manipulative Techniques: AI that deploys subliminal techniques beyond a person's consciousness, or purposefully manipulative techniques exploiting vulnerabilities
- Real-Time Remote Biometric Identification: Use in publicly accessible spaces for law enforcement, with narrow exceptions requiring judicial authorization
- Emotion Recognition in Workplaces and Schools: AI systems that infer emotions of individuals in workplace or educational settings
- Untargeted Facial Image Scraping: Building facial recognition databases through untargeted scraping from the internet or CCTV
- Predictive Policing (Individual): AI systems making risk assessments of natural persons to predict criminal offenses based solely on profiling
Limited Exceptions
The Act provides narrow exceptions for law enforcement use of real-time biometric identification, subject to strict conditions: prior judicial authorization, necessity for specific serious crimes, and proportionality assessment. These exceptions are tightly circumscribed and require national implementation.
Read our analysis of the February 2025 enforcement milestone →