The European Union has officially approved the AI Act, a landmark regulation that will prohibit certain uses of AI and demand transparency from providers. The act has been hotly debated and will take years for some rules to be enforced. It divides its rules based on the level of risk an AI system poses to society, with stricter rules for higher-risk systems. The compromise on the AI Act may not have an immediate effect on major AI players like OpenAI and Google, as it likely won’t take effect until 2025. Some member states, such as France and Germany, lobbied to water down restrictions on general-purpose AI systems during negotiations. The act also sidesteps controversial issues around generative AI and does not create new laws around data collection. It won’t apply stiff fines to open-source developers and smaller companies, and it may pressure other political figures, particularly American policymakers, to move faster. The US, in contrast, has largely failed to get AI regulation off the ground. The AI Act demonstrates where the EU stands on AI and may influence other countries’ approaches to AI regulation.
