The European Union is preparing guidance on compliance with its new law on artificial intelligence. As part of that process, it's launching a consultation here, seeking input in two areas: first, how the law defines AI systems as opposed to "simpler traditional" software. It's in effect a call to people in the AI industry, business, academia, and civil society for views on the clarity of key elements of the Act's definition and examples of software that should be out of scope.

The second demand is for prohibited uses of AI. The Bill contains a few use cases that are prohibited as they meet the "unacceptable risk" definition, such as the social scoring of the Chinese model. Most of the consultation is concentrated here. The EU wants to be given highly detailed comments on each one of the prohibited uses and is particularly keen on examples of practical application.

The public consultation is open until 11 December 2024. The Commission will publish guidelines on the definition of AI systems and prohibited uses in the first quarter of 2025, according to the press release.