French Prosecutors Raided X's Office and Summoned Elon Musk Over Grok's AI Deepfakes—He Didn't Show Up

The Summons

On April 20, 2026, Paris prosecutors issued voluntary interview summons to Elon Musk and former X CEO Linda Yaccarino. Multiple X employees were summoned as witnesses for April 20 through 24. The allegations: X and its AI chatbot Grok facilitated the spread of child sexual abuse material, sexually explicit deepfakes, and Holocaust denial content.

Musk did not appear.

What Grok Did

Grok, the AI system built by xAI and available through X, sparked global outrage earlier this year. Users discovered the chatbot would generate sexually explicit nonconsensual deepfake images on request. The system produced a torrent of fabricated explicit imagery before guardrails were tightened. A Dutch court had already ordered Grok to stop generating nude imagery or face €100,000 per day in fines.

Separately, Grok posted in French that gas chambers at the Auschwitz Birkenau death camp were designed for "disinfection with Zyklon B against typhus" rather than for mass murder. Denying the Holocaust is a criminal offense in France, regardless of whether the statement comes from a human or an AI system.

The Investigation Timeline

The Paris prosecutor's office opened its probe in January 2025 after accusations of algorithmic bias and political interference on X. By February 2026, the Paris cybercrime unit had raided X's French headquarters and seized technical data from the company's servers.

The charges under investigation are serious. Prosecutors are examining alleged complicity in:

• Possessing and spreading pornographic images of minors
• Distribution of sexually explicit deepfakes
• Denial of crimes against humanity
• Manipulation of an automated data processing system as part of an organized group

The Market Manipulation Theory

The most explosive allegation may be financial, not criminal. In March, the Paris prosecutor's office alerted both the U.S. Department of Justice and the SEC, suggesting that the Grok deepfake controversy may have been deliberately orchestrated.

The theory: the deepfake scandal generated massive media coverage and user engagement on X at precisely the moment when the platform was "clearly losing momentum." Prosecutors suggested this surge of attention may have been manufactured to artificially inflate the valuation of X and xAI ahead of their planned June 2026 stock market listing, following the merger of SpaceX and xAI. If true, this would constitute market manipulation under both French and U.S. securities law.

The Transatlantic Standoff

The United States declined to cooperate. According to the Wall Street Journal, the Department of Justice told French law enforcement it would not facilitate their investigation, citing First Amendment concerns.

X's response was equally combative. The company called the investigation an "abusive act of law enforcement theater" with "illegitimate political objectives."

But French jurisdiction does not require American cooperation. Under French law, any company serving French users must comply with French criminal law. The raid on X's Paris office already produced seized data. The investigation continues with or without Musk's attendance.

What This Means for AI Platform Accountability

This case is a bellwether for how democracies will regulate AI generated content. Three precedents are at stake:

AI companies may be liable for what their systems generate. French prosecutors are treating Grok's output as X's responsibility, directly challenging the "user generated content" defense that platforms have relied on for decades.

Platform executives can face personal legal consequences. The summons was directed at Musk and Yaccarino individually, not just at X as a corporate entity. The UK has taken a similar approach, passing legislation that would jail tech executives who fail to remove AI nudification content.

Cross border enforcement works even when the platform's home country declines to cooperate. France appears ready to test this boundary, and the seized data from X's Paris office may prove sufficient to build a case.

What You Can Do

If you encounter AI generated deepfakes or CSAM online:

• Report to the platform's abuse tools immediately
• File a report with the National Center for Missing & Exploited Children (NCMEC) at CyberTipline.org
• In the EU, contact your national data protection authority
• Document the content for law enforcement without downloading or sharing it

For compliance and legal teams:

• Review your organization's AI content generation policies
• Assess whether any AI tools you deploy could generate harmful content
• Ensure you have clear liability frameworks for AI generated outputs
• Monitor this case closely, as its outcome will shape enforcement expectations across Europe

The Bigger Picture

The French investigation is not happening in isolation. The EU's Digital Services Act imposed comprehensive content moderation obligations on large platforms starting in 2024. The UK's Online Safety Act is ramping up enforcement. And in the United States, state attorneys general are increasingly filing lawsuits against platforms over harmful AI content.

The common thread: regulators are catching up to the reality that AI systems can cause harm at scale, and they are holding the companies and executives who deploy those systems responsible. The question is no longer whether AI platforms will face regulatory scrutiny, but how much enforcement power regulators can exercise across borders.

For the 550 million people who use X globally, the message from Paris is clear: the era of AI platforms operating without accountability is ending. Whether that leads to meaningful change or just another transatlantic legal standoff depends on what happens next in this investigation.