As of 2 February 2025, the EU AI Act has introduced rules regarding prohibited artificial intelligence (AI) practices and AI literacy requirements, with additional enforcement mechanisms set to take effect in six months.
The Act bans several AI practices, including:
- Facial recognition databases created via online scraping or security footage.
- Biometric identification systems.
- Manipulative AI techniques aimed at influencing behaviour.
- Social scoring systems.
- Criminal prediction software (law enforcement agencies may still use real-time facial recognition in public spaces or criminal investigations).
- Emotional detection AI in workplaces and schools.
- AI systems that exploit vulnerabilities based on age, disability, or socioeconomic status.
The European Commission has issued nonbinding guidelines to clarify these prohibitions and ensure ethical AI compliance. A final draft of the Code of Practice for General-Purpose AI Models is expected in April.
Companies violating the AI Act may face fines of up to 7% of annual revenue or €35 million, whichever is greater.
AI literacy is increasingly seen as a key concept in AI governance.
