The draft aims to provide a uniform definition for AI and impose obligations on the providers and users with a risk-based approach. The draft aims to eliminate the possible discriminatory or manipulative uses of AI systems that may be harmful to people’s fundamental rights or environment. By this mean, a new set of rules to prevent AI systems from social scoring, biometric categorization and emotion recognition will be introduced.

On the other hand, it is also addressed that generative AI systems, such as GPT, contain high risks for the breach of fundamental and environmental rights as well as democracy and rule of law, so the draft also aims to ensure these systems fully fulfill transparency requirement and take necessary measures from generating illegal content.

Finally, in the draft Act the citizens are provided the right to file complaints about AI systems.

Sources:

European Parliament. (May 11, 2023). “AI Act: a step closer to the first rules on Artificial Intelligence”. Retrieved June 18, 2023, from https://www.europarl.europa.eu/news/en/press-room/20230505IPR84904/ai-act-a-step-closer-to-the-first-rules-on-artificial-intelligence European Parliament. (June 14, 2023). “MEPs ready to negotiate frst-ever rules for safe and transparent AI”. Retrieved June 18, 2023, from https://www.europarl.europa.eu/news/en/press-room/20230609IPR96212/meps-ready-to-negotiate-first-ever-rules-for-safe-and-transparent-ai