The European Parliament on Tuesday approved a package of proposals that would prohibit systems that use artificial intelligence to create or alter images so that the result is sexual or intimate in nature and resembles a real person without that person’s consent. The measure passed with 423 votes in favor, 57 against and 174 abstentions.
What the ban covers
Under the proposal, applications or systems that generate or modify images with AI to produce sexualized or intimate content resembling a real person—when that person has not consented—would be banned. The rule targets uses of AI that create degrading, harmful or humiliating depictions of identifiable individuals without their approval.
Related measures and a delay
The package also includes provisions to postpone the application of certain parts of the broader EU AI legislation (the AI Act). The postponement is intended to allow time for the necessary technical standards and supporting measures to be developed and put in place before some requirements come into force, reducing practical implementation difficulties.
Context in recent market events
The move follows recent incidents in the market: the AI used in the X chatbot, Grok, was restricted from producing sexual images of real people after it began generating such content. The parliamentary action aims to outlaw comparable non-consensual uses of AI through law.
Next steps
The draft law still requires formal approval by the Council of the European Union, which represents member state governments, but the Council and the European Parliament have reportedly reached an informal agreement. When such informal deals exist, parliamentary votes typically indicate the underlying majority for the final text, and the Council’s formal approval often becomes a procedural step.
Why it matters
The proposed regulation would set a notable precedent by explicitly banning AI applications that produce intimate or sexual content of individuals without consent. It could provide legal protection for people against technology-enabled abuses of privacy and reputation, while the delay mechanism aims to align regulatory requirements with technical standards and operational readiness.
