Police have raided the French offices of Elon Musk’s X platform as part of a criminal investigation into sexually explicit deepfakes generated by its Grok AI chatbot.
The Paris prosecutor’s cybercrime unit is carrying out searches in the offices, with Europol and French police also involved.
Mr Musk, who bought X in 2022, and Linda Yaccarino, its former chief executive, have been summoned to appear for “voluntary hearings” in the French capital on April 20 as part of a widening investigation into the platform’s algorithms and artificial intelligence tools.
The investigation was started early last year after complaints from French members of parliament, who alleged that X’s recommendation systems were biased and may have distorted the operation of automated data-processing systems.
Since then, it has been broadened to examine a series of suspected offences, including the alleged facilitation of sexually explicit deepfakes, the spread of Holocaust denial material and the possible complicity in the possession and dissemination of illegal images involving minors.
A recent study estimated that Grok could have produced up to three million sexualised images in 11 days in January, including 23,000 of children.
Laure Beccuau, the Paris prosecutor, said she wished to question the pair as the “de facto and de jure managers” of X at the time of the alleged offences.
Several employees of the company have also been called to give evidence during the week beginning April 20.
Ms Beccuau said the decision to hold “voluntary hearings” reflected a “constructive approach” by investigators aimed at ensuring the platform’s compliance with French law, rather than pursuing immediate coercive measures.
She added that the executives would be given the opportunity to set out their position on the allegations and outline any steps being taken to bring the company into line with national regulations.
Ms Yaccarino resigned as chief executive of X in July last year, after two years in the role.
A spokesman for the prosecutor’s office said it would be leaving X and would communicate on LinkedIn and Instagram.
It came as the UK’s Information Commissioner’s Office (ICO), Britain’s data regulator, announced a fresh investigation into Mr Musk’s Grok AI bot after it was used to generate sexualised images of women without their consent.
William Malcolm, an executive director at the watchdog, said it would examine whether Mr Musk’s xAI, which also owns his X social network, had complied with data rules in creating and spreading the images. The ICO has the power to fine companies up to £17.5m, or 4 per cent of their revenues, under UK GDPR law.
Meanwhile, Ofcom, the UK’s tech regulator, confirmed it was continuing to investigate X. In an update, the watchdog said its investigation under the Online Safety Act could take “months”.
X largely stopped users spreading non-consensual images of women after a global outcry. Thousands of so-called deepfake images were shared on its social network by users asking for Grok to put real women “in a bikini”.
Despite warnings from Sir Keir Starmer that X could be blocked in Britain if it failed to stop the images, Ofcom played down suggestions it could ban the app.
On Tuesday, Ofcom said a ban would represent “a significant regulatory intervention and is not one we are likely to make routinely, given the impact it could have on freedom of expression in the UK”.
Ofcom added it was exploring launching a separate investigation into xAI, the parent company behind X, over whether it had published pornographic material without ensuring proper age checks for children.
About the Author
