6 things to know about the EU AI Act
The regulation of artificial intelligence in the European Union will be governed by the AI Act, marking the world's inaugural comprehensive legislation on AI
On December 8, the European Parliament and Council reached a provisional agreement on the Artificial Intelligence Act.
Here are six key points:
This legislation aims to safeguard fundamental rights, democracy, and environmental sustainability from high-risk AI applications.
It introduces bans on specific AI uses, such as biometric categorization, social scoring, and manipulation of human behavior. Law enforcement use of biometric identification systems is subject to strict conditions.
High-risk AI systems face mandatory impact assessments, and citizens have the right to launch complaints and receive explanations.
The Act includes measures to support innovation (including the ability of SMEs to access technology.)
The Act brings teeth to its enforcement measures, prescribing fines ranging from 35 million euros or 7% of global turnover to 7.5 million euros or 1.5% of turnover for non-compliance.
The agreement awaits formal adoption by both Parliament and Council to become EU law (which would take effect in 2025 at the earliest.)
I have a key observation. The EU AI Act has key differences to Biden’s “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence”.
The EU guidelines appear to be more focused on nefarious activity mitigation as well as retribution, while Biden’s Executive Order offers more insights on job creation and commercialization.