A European Union-wide ban on AI systems with “unacceptable” risk came into force on Feb. 2 as the first provisions of the EU’s AI Act took effect. Problems persist, however, over what the legislation requires and what corporate practices or uses of data may risk flouting the rules.



