The European Union is developing new legal compliance guidance regarding AI. As part of that process, we have launched a council (here) seeking input on two areas. The first is how the law defines AI systems (as opposed to “simpler traditional” software). We are asking people from the AI industry, business, academia, and civil society to comment on clarity on key elements of the law's definition, as well as examples of software that should be excluded.
The second question concerns prohibited uses of AI. The law includes several use cases, such as Chinese-style social scoring, that are considered “unacceptable risks” and are prohibited. Most of the consultation will be focused here. The EU wants detailed feedback on each prohibited use and is particularly keen on practical examples.
The consultation will run until December 11, 2024. The European Commission will then issue guidance on the definition and prohibited uses of AI systems in early 2025, according to a press release.