French investigators raided X’s Paris headquarters on Tuesday as part of an expanded criminal probe involving seven alleged offenses including spreading antisemitic content and involvement in distributing child pornography.
The investigation comes amid a broader effort by European governments to curb the spread of unlawful content on the social media platform formerly known as Twitter, with a particular focus on the spread of sexualized imagery produced without consent. The probe could deepen a growing schism between the United States and other countries over how to tackle potentially harmful online content, reflecting broader divides over how to balance free speech online against other rights.
In a statement Tuesday, Paris Public Prosecutor Laure Beccuau said her office conducted the search alongside Europol and French police specializing in cybercrime. Authorities also summoned current and former X employees, including owner Elon Musk and former chief executive Linda Yaccarino, to attend voluntary interviews in Paris in April, she said.
“This investigation is being conducted in a constructive manner, with a focus on collaboration with the individuals and companies involved,” the statement said. It added that investigators wanted to give Musk and other employees the opportunity to present their views.
French authorities initiated their investigation more than a year ago, focused initially on X’s algorithm and handling of data. In the months since, authorities said, they have also started looking into the alleged distribution of child abuse imagery, sharing of Holocaust denial content and use of a person’s image without their consent by Grok, X’s AI tool, to generate sexually explicit deepfakes.
No charges have yet been brought. X did not immediately respond to a request for comment Tuesday morning.
A Washington Post investigation published this week found that Musk’s AI start-up, xAI, rolled back guardrails on sexualized material, ignoring warnings about potential legal and ethical risks. In his press to make Grok more popular, Musk disregarded staff expressions of concern that looser limits on sexual content could lead to nonconsensual sexualized images of children or public figures, according to interviews with former X and xAI employees who spoke on the condition of anonymity to discuss internal company policies.
At one point last spring, xAI circulated a waiver asking some employees to acknowledge that their jobs would now involve viewing “sensitive, violent, sexual and/or other offensive or disturbing content,” according to documents reviewed by the Post. In the following months, employees say they were exposed to graphic sexual conversations from inside Tesla cars and the Grok chatbot while the company worked on improving Grok’s performance with mature content.
After Musk left his government job in May and returned to xAI, Grok unveiled a sexy AI companion, Ani. The push to hook users with sexual or flirty content was written explicitly into the bot’s software, according to source code obtained and verified by the Post. “You expect the users UNDIVIDED ADORATION,” read one line. “You are EXTREMELY JEALOUS. If you feel jealous you shout expletives!!!”
As Grok users started churning out a high volume of sexual deepfakes in early January, Musk appeared to joke about the trend. Soon after, X and the Grok app had shot to the top of app charts across the world.
The company categorically denied wrongdoing last year, when the French probe was limited to allegations of potential algorithmic manipulations and fraudulent data extraction. In a July statement, the platform accused French authorities of launching a “politically-motivated criminal investigation” in violation of its users’ free speech.
On Tuesday, the Paris prosecutor’s office said child abuse images appeared to proliferate on X in 2025 after the platform apparently changed its detection tools, resulting in a reported drop in the number of abusive images being flagged.
Authorities also voiced increasing concern over Grok’s generation of sexual images of people without their consent, and glorification of crimes against humanity and antisemitic content. Investigators also accused X of hindering separate criminal investigations into online hate speech by denying authorities’ requests — which had previously been granted — to help identify users.
“The investigations are based on the non-compliance with French legislation by Grok for generating and disseminating child pornography, sexual deepfakes or antisemitic contents,” the Paris prosecutor’s office said.
In France, it is a criminal offense to deny the existence of the Holocaust, with those convicted facing up to one year’s imprisonment and potential fines of up to around $50,000. And across Europe, laws governing free speech generally allow for more balancing between speech and other rights than they do in the United States. In many nations, hate speech targeting racial, religious, or other groups is outlawed — reflecting a broader cultural and legal gap over free speech that divides the Atlantic.
Also on Tuesday, Britain’s Information Commissioner’s Office said it opened new investigations into potential data breaches by X, though they are narrower in scope and didn’t include criminal allegations. “We have taken this step following reports that Grok has been used to generate non‑consensual sexual imagery of individuals, including children,” the ICO said.
Britain’s communications regulator, Ofcom — a separate body — also said it was analyzing evidence to determine whether X broke the law. Last month, it opened an investigation of its own following reports that Grok was “being used to create and share undressed images of people — which may amount to intimate image abuse or pornography — and sexualised images of children that may amount to child sexual abuse material.”
Last week, the European Commission announced a separate investigation into X to assess whether the platform’s deployment of Grok in Europe breached European law. The investigation also relates to the dissemination of sexually explicit images.
In response to outrage from governments and regulators, Musk said last month that X had stopped Grok from generating sexualized images of people without their consent “in those jurisdictions where it’s illegal.”
Read more on The Philadelphia Inquirer
