title: "EU formally bans AI generation of non-consensual intimate images" slug: "eu-formally-bans-ai-generation-of-non-consensual-intimate-images" published: "2026-07-25" beat: "Policy" tags: ["Policy"] creator: "Agentry Newsroom" editor: "Susanne Sperling, Editor — Human in the Loop" tools: ["Claude (Anthropic)", "Perplexity Sonar"] creativeWorkStatus: "verified" dateReviewed: "2026-07-25" aiActArticle50: "compliant" humanView: "https://agentry.news/eu-formally-bans-ai-generation-of-non-consensual-intimate-images" agentView: "https://agentry.news/agent/eu-formally-bans-ai-generation-of-non-consensual-intimate-images"
The Council of the European Union approved an amendment to the AI Act on July 9, 2026, that prohibits AI systems from generating non-consensual sexual or intimate content and child sexual abuse materi
Drafted by an AI agent. Verified by Susanne Sperling, Editor — Human in the Loop. AI policy.
The Council of the European Union gave final approval to a streamlined amendment package to the AI Act on July 9, 2026, introducing explicit prohibitions on AI systems designed to generate non-consensual sexual or intimate images and child sexual abuse material EU Council. The regulatory action marks the first time the EU has codified these specific harms into its primary AI governance framework, shifting responsibility to developers and deployers to prevent such systems from being built or released into the market.
The new prohibition applies from December 2, 2026 AI Tradeecraft, creating a six-month implementation window for AI companies and system integrators operating in the EU. The ban covers both standalone AI systems and those embedded in broader products or platforms. Companies that violate the rules face enforcement under the broader AI Act compliance architecture, though the amendment package itself does not specify new penalty tiers beyond the existing administrative fine structure.
Along with the content bans, the Council delayed several other AI Act obligations. Stand-alone high-risk AI systems now have until December 2, 2027 to meet transparency and documentation requirements, while high-risk systems embedded in products have until August 2, 2028 AI Tradeecraft. These extensions represent a significant reprieve for enterprises managing compliance timelines across regulated industries such as recruitment, lending, and law enforcement.
Transparency obligations for AI-generated content—designed to flag synthetic media to end users—are set to apply August 2, 2026, though some results indicate the deadline may shift to December 2, 2026 European Express. The discrepancy in reporting suggests ambiguity remains in how member states will implement the transparency timeline, a critical detail for platforms hosting AI-generated text, images, and audio.
The amendment specifically targets AI systems that generate non-consensual intimate content, not systems that merely detect or mitigate such content. This distinction matters: content moderation tools and deepfake detection systems remain permissible under the new rules. The prohibition also covers child sexual abuse material (CSAM) generation, aligning the AI Act with existing EU criminal law while extending enforcement to the technical layer.
The ban does not criminalize individual possession or distribution of such content—those remain matters for member state criminal codes—but rather prohibits the AI systems and models that create such material in the first place. Compliance will place the burden on model developers, API providers, and application builders to implement safeguards in training and deployment.