EU bans sexualized deepfake AI tools, effective Dec 2026

EU member states have approved new legislation that forbids the use of artificial‑intelligence applications to create sexualized deepfake content. The ban, which also covers depictions of child sexual abuse, will come into force on 2 December 2026 and will be enforced by the EU AI Agency.

The law targets both the users of such tools and the providers of the underlying software, expanding existing rules that focus only on the perpetrators. Satirical caricatures are exempt, and non‑consensual bikini images created by certain tools may remain permissible. The regulation also foresees mandatory AI‑generated content watermarks, which providers such as ChatGPT, Claude and similar services must implement by the same deadline.

By requiring clear labeling of AI‑generated media and restricting the creation of harmful sexualised imagery, the EU aims to curb digital gender‑based violence and protect minors online.