The European Council agreed on a proposal to streamline Artificial Intelligence Act rules as part of the “Omnibus VII” legislative package, which is intended to simplify the European Union’s digital legislative framework and harmonize Artificial Intelligence regulations. The proposed amendments include a prohibition on generating non-consensual sexual content, a fixed timeline for applying high-risk Artificial Intelligence rules, and obligations for Artificial Intelligence system registration in the European Union database.
The proposal also postpones the establishment of Artificial Intelligence regulatory sandboxes until December 2027 and clarifies the Artificial Intelligence Office’s supervisory powers. Following the Council’s approval, negotiations with the European Parliament will commence. The changes align with the European Council’s broader push for a simplified regulatory framework designed to boost competitiveness and reduce burdens on businesses, particularly SMEs.
In parallel, the European Union is reportedly considering banning Artificial Intelligence-powered nudification tools that can create non-consensual explicit images of women and minors. The move follows backlash against X’s Grok and its “spicy mode,” which allowed users to generate sexualized deepfakes. A proposal set to be approved by European Union ambassadors on Friday would make it illegal to market in EU any Artificial Intelligence system that can generate non-consensual sexualized videos, images or audio files involving real people.
The European Commission has also been investigating X and Grok for potential violations of the Digital Services Act and is considering whether such Artificial Intelligence-generated sexual images should be classified as unacceptable risks under the Artificial Intelligence Act. Pressure for action has intensified, with over 100 organizations, including Amnesty International and Interpol, urging faster intervention and warning that these tools pose a significant threat to child safety and human dignity.
