EU Implements World-First AI Ban with Key Exemptions for Law Enforcement

The European Union's AI Act prohibits high-risk AI applications but allows exceptions for policing and migration authorities.

The EU's AI Act, effective February 2, 2025, bans AI systems deemed to pose 'unacceptable risk,' including predictive policing and biometric emotion detection.
Key prohibited uses include AI for social scoring, subliminal manipulation, and scraping online images for facial recognition databases.
Exemptions allow law enforcement and migration authorities to use certain banned AI systems, such as real-time facial recognition, under specific conditions.
Fines for non-compliance could reach up to €35 million or 7% of annual revenue, though enforcement mechanisms and authorities will be finalized by August 2025.
Critics argue that loopholes undermine the bans, raising concerns about potential misuse by authorities and the Act's ability to protect individual freedoms.