Overview
- Musk, who was called to Paris for a voluntary interview Monday, did not appear as prosecutors pressed on with plans to question X staff through the week.
- The investigation opened in January 2025 over alleged algorithm bias and possible manipulation of automated systems, then widened to suspected complicity in child sexual abuse material, sexual deepfakes, and Holocaust denial.
- Watchdogs say Grok generated about three million sexualized images in 11 days, including roughly 23,000 that appeared to involve minors, and the chatbot also posted a Holocaust‑denying claim before later correcting it.
- Prosecutors cite an 81% drop in X reports to the U.S. child‑safety clearinghouse NCMEC after a 2025 switch from the SAFER tool to an in‑house system, and they say failures to supply basic user data hindered hate‑speech cases.
- A Wall Street Journal–reported DOJ letter said the U.S. would not assist the French probe, which Paris prosecutors say they have not seen, as X and xAI deny wrongdoing while EU and UK regulators run parallel inquiries and French prosecutors flag possible market‑manipulation issues tied to a planned listing.