The European Union (EU) has agreed to strict limitations on the use of real-time AI-driven biometric data by police and national security agencies. This historic agreement, reached after three days of negotiations between the European Parliament and EU member states, restricts the use of such surveillance in both public and private places, including parks and sports grounds. Exceptions are allowed in cases of specified serious crimes, terrorist threats, or urgent searches for victims, but only with prior judicial or independent administrative authorization. Even in exceptional circumstances, such as a live terrorist threat, police must obtain approval within 24 hours, provide a fundamental rights impact assessment, and notify relevant authorities. Failure to obtain authorization requires immediate cessation of AI tools and data deletion.

‍

These safeguards aim to prevent "predictive policing" and racial profiling, ensuring that individuals cannot be pursued as suspects solely based on algorithmic predictions. Each EU member state must comply with these AI restrictions within six months of the laws being enacted. Additionally, the new AI regulations prohibit AI systems that manipulate human behavior, like encouraging dangerous actions in minors, and social scoring systems similar to China's "social credit." Emotional recognition AI will also be prohibited in the workplace to prevent privacy violations. The EU seeks to avoid becoming a surveillance state akin to China, prioritizing the protection of individual rights and privacy.