European Union officials reach an agreement to ban “sexualised” AI deepfakes, according to reporting from multiple outlets. The prohibition is set to be incorporated into amendments to the EU’s broader framework for artificial intelligence, which was adopted in 2024. The deal is part of the ongoing process of refining the EU’s AI regulations and clarifying what types of AI content and systems are restricted.
Separately, officials also agree to adjust parts of the timetable for implementing other provisions. Reporting indicates there is a delay for the rollout of rules covering “high-risk” AI systems—models and applications considered potentially dangerous to safety, health, or citizens’ fundamental rights. The high-risk category is tied to obligations intended to reduce harm from AI used in sensitive contexts.
Together, the reports describe a package of changes that combines a new ban focused on sexualised AI-generated deepfakes with a shift in implementation timing for rules aimed at high-risk AI systems.
