The European Commission opened an investigation on Monday into X (formerly Twitter) and the use of its AI chatbot Grok to generate sexually explicit images of women and underage girls without their consent.
The investigation will determine if the AI tool violates the Digital Services Act (DSA) and whether X failed its obligation to properly assess and mitigate the risks of deploying Grok into the EU. The platform’s after-image editing capabilities allow users to enter a prompt under any photo and have the AI create realistic changes to the original, referred to as deepfakes.
X is classified as a Very Large Online Platform (VLOP) under EU law and thus bears the responsibility of ensuring that its services do not infringe on the rights of EU citizens and do not disseminate illegal content, as per articles 34 and 35 of the DSA.
If X does not adequately adjust its producers during the investigation they may face interim restrictions and have already faced a €120 million fine related to noncompliance during the 2023 investigation. The commission will also extend its ongoing 2023 investigation concerning X’s broader risk management obligations.
The investigation marks the latest development in the crackdown against AI deepfakes, which has already occurred in numerous jurisdictions, including California, Australia, India, the UK, and China.
Earlier this month, X implemented measures to prevent users from generating explicit content through Grok following immense backlash. However, measures have been imperfect, and investigations remain ongoing.