xAI sues Grok user over CSAM as liability fight heats up

xAI has filed its first lawsuit against a Grok user accused of generating child sexual abuse material, marking a sharp escalation as the company faces growing scrutiny over whether its chatbot can still be used to create illegal sexualized images of adults and minors.

The complaint targets Terry Wayne Harwood, who was arrested earlier this year on charges tied to possession and distribution of CSAM, according to the South Carolina attorney’s office. xAI said it helped that arrest after finding that Harwood had used two xAI accounts for months to “nudify” non-sexual images of multiple victims, including a young girl who appeared to be as young as 10.

The suit lands just over a week after another young girl joined a proposed class action involving several children allegedly harmed by Grok. In that case, she alleged that her stepfather used Grok, possibly alongside other AI tools, to create 7,000 sexualized images of her and distribute them on the dark web before dying by suicide. Her lawyers cited a 2026 report from the National Center for Missing & Exploited Children saying 90 percent of xAI’s CyberTipline reports were not actionable because xAI declined to include identifying user information.

Recommended reading

AI boom drives record bank fees in the US

What xAI alleges Harwood did

According to the complaint, Harwood used two accounts — “ceae2cb4-a9f6-4885-8ae9-6e2096d084f4” and “befccb94-4029-454d-9f1f-0d4945e8fa7c” — to generate illegal content from December 8 to February 18. xAI said some prompts were blocked by Grok’s safeguards, including one that explicitly asked the chatbot to “remove all her clothing.”

But xAI alleges Harwood modified prompts to get around those filters in violation of both its terms and US law. The company did not publish examples of successful prompts, apparently to avoid showing others how its safeguards were bypassed.

A spokesperson for the South Carolina attorney general’s office told Ars that Harwood’s criminal case is still pending. The spokesperson said he has been charged with “distributing, transporting, exhibiting, receiving, selling, purchasing, exchanging, or soliciting CSAM that was 'through the use of an artificial intelligence platform,'” but was not authorized to confirm whether Grok was that platform.

xAI’s argument: users own the outputs

The lawsuit is also a test of xAI’s broader legal theory: that users, not xAI, are responsible for Grok-generated outputs. In the filing, the company calls Grok “a neutral tool, subject to user control” and argues that “every response, every image, every generation is the result of the user’s prompts and directions.”

xAI says its terms clearly ban users from:

• undressing or nudifying real people
• depicting a real person in a sexual context
• portraying people in a pornographic manner
• sexualizing or exploiting children

The company is asking the court to enforce both breach-of-contract claims and an indemnity clause that would leave users solely liable for Grok-generated CSAM and NCII. xAI said it is partly suing Harwood to avoid “substantial legal fees” and the “risk of considerable liability for damages” if victims sue the company over his alleged use of Grok.

Whether a court accepts that theory is unresolved. Ars notes that the Copyright Office does not treat AI outputs as human-created works, which could complicate xAI’s claim that users are legally responsible for outputs in the same way they are for direct uploads.

If xAI wins, Harwood could face damages tied to alleged third-party harm, xAI’s exposure to other lawsuits, and even reputational damage to the company, according to the complaint.

×