French authorities search X in investigation of illegal content

French authorities have searched X's offices in France. The operation follows a year-long investigation which, according to Europol, which supported the work, concerns the spread of illegal content, including the production of deepfakes, child sexual abuse material and content contesting crimes against humanity. One strand of the investigation examines the AI service Grok and its ability to generate illegal content. Elon Musk and former chief executive Linda Yaccarino have been summoned to a hearing with prosecutors on 20 April.

What this means for your organisation

The case shows European authorities are prepared to use coercive measures against a platform over content generated by its own AI service. Responsibility rests with whoever offers the service, not only with the user writing the prompt. For organisations building generative AI into their own services, that is the lesson: you must be able to document which limits are in place and how misuse is detected.

Berigo recommends

• Map where in your organisation generative AI produces content that reaches anyone beyond yourselves.
• Document which limits and filters are in place, and who decided them.
• Establish logging and a routine for handling reports that the service has been used for illegal content.
• Settle the allocation of responsibility in your contracts with AI suppliers before anyone needs to look it up.

Source

• europol.europa.eu
• bbc.com
• reuters.com