- EU finalizes historic AI regulations.
- High-risk AI under strict oversight.
- Bans on sensitive biometrics, facial image scraping, emotion recognition, and AI manipulation.
European Union policymakers and lawmakers reached a consensus on Friday to establish the world’s first comprehensive set of regulations governing the use of artificial intelligence (AI). The rules, applicable to a wide array of AI applications, including technologies like ChatGPT and biometric surveillance tools, are set to come into effect early next year and be fully implemented by 2026.
Here are the key highlights of the groundbreaking agreement:
1. High-Risk AI Systems:
- AI systems with the potential to significantly harm health, safety, fundamental rights, environment, democracy, elections, and the rule of law are classified as high-risk. They must comply with stringent requirements, including undergoing a fundamental rights impact assessment before gaining access to the EU market.
2. Transparency for Limited-Risk AI:
- AI systems posing limited risks will face lighter transparency obligations. This includes the use of disclosure labels indicating that content is AI-generated, allowing users to make informed decisions.
3. AI in Law Enforcement:
- Real-time remote biometric identification systems by law enforcement in public spaces are limited to specific scenarios, such as identifying victims of kidnapping, human trafficking, sexual exploitation, and preventing present terrorist threats.
4. GPAI and Foundation Models:
- General Purpose AI Systems (GPAI) and foundation models will be subject to transparency requirements, technical documentation, compliance with EU copyright law, and dissemination of detailed summaries about content used for algorithm training.
5. Prohibited AI:
- The regulations explicitly prohibit various AI applications, including biometric categorization using sensitive characteristics, untargeted scraping of facial images, emotion recognition in workplaces and educational institutions, social scoring based on personal characteristics, and AI manipulation of human behavior to circumvent free will.
6. Sanctions for Violations:
- Fines for AI regulation violations will range from 7.5 million euros or 1.5% of global annual turnover for smaller infringements to as much as 35 million euros or 7% of global turnover for more severe breaches.
The agreement encourages companies to voluntarily commit to the AI Pact, implementing key obligations even before the official legislation takes effect. While the details are expected to be finalized in the coming weeks, the agreement marks a historic step in establishing ethical and accountable practices in the rapidly advancing field of artificial intelligence.
[embedpost slug=”/samsung-galaxy-a33-new-price-in-uae-specs-dec-2023/”]