The European Union has reached an agreement to ban AI systems capable of generating sexualized deepfakes, marking a firm boundary against the use of artificial intelligence for humiliation and exploitation. The deal was struck between EU lawmakers and member countries, as reported in May 2026.
The agreement draws a clear line on one of the more harmful applications of generative AI, targeting systems specifically designed to produce non-consensual sexualized imagery of individuals.
Alongside the deepfake prohibition, the EU has decided to delay the implementation of rules governing high-risk AI systems. Those requirements are now scheduled to take effect in late 2027 and 2028. The delay is intended to give businesses more time to comply and to support broader innovation within the bloc.
The EU is also moving to address cybersecurity risks posed by advanced AI models, though specific details of those measures were not disclosed in the source material.
The agreement could signal a broader shift in how the EU enforces its AI Act, balancing consumer protections against the pace of technological development. By banning sexualized deepfake generation outright while easing the timeline for high-risk AI compliance, the bloc appears to be distinguishing between harms it considers immediately unacceptable and those requiring more time to regulate effectively.
Source: Tech-Economic Times