The European Parliament's civil liberties committee approved draft text for the AI Omnibus law. This legislation includes an amendment to outlaw AI systems that generate realistic, nonconsensual sexually explicit images. Alignment with European governments makes the ban likely to be adopted into law later this year.
AI developers must implement direct safeguards on their models to comply with the new rules. The approved text delays implementation for high-risk AI applications until late 2027. Additional compliance requirements will not take effect until 2028.
This extension provides companies more time to prepare for complex regulatory demands. The move signals a significant step in global generative AI regulation. These rules directly impact major US tech firms operating in the region.