Grok user tried to direct child porn movie, now xAI is seeking “reputational damages”
Grok users are still trying to create child porn
Image by Getty/Nur Photo
- xAI is suing Terry Wayner Harwood for allegedly using Grok to create non-consensual sexual deepfakes.
- Harwood allegedly targeted women, girls, and a child estimated at 10 to 11 years old, despite Grok refusing explicit prompts.
- Investigators arrested Harwood in March 2026 and charged him with eight counts of child sexual exploitation.
- The lawsuit follows backlash over Grok deepfakes, prompting xAI to add safeguards and warn users against illegal content.
Key Takeaways by nexos.ai, reviewed by Cybernews staff.
Elon Musk’s xAI, now renamed to SpaceXAI, is suing a South Carolina man for violating its updated rules following Grok’s deepfake pornography scandal.
Terry Wayner Harwood, 67, of South Carolina, repeatedly and deliberately violated xAI's terms of service and acceptable use policy through his “abhorrent abuse of…generative AI chatbot known as Grok,” the lawsuit alleges.
Harwood allegedly created at least two accounts using fake identities where he uploaded clothed images of women and girls without their consent.
He then instructed Grok to undress the victims and put them in pornographic positions.
Using one of these accounts, Harwood asked Grok to create sexual images of one victim who is estimated to be between 10 and 11 years old.
The prompt asked Grok to generate a video where the child was artificially manipulated into doing “her best Playboy model impression.”
Harwood asked Grok to create deeply disturbing explicit material – something the chatbot refused to do.
That didn’t stop him from submitting “additional modified prompts” to jailbreak the AI model and get around its safeguards, according to the lawsuit.
Harwood's prompt provides insight into his motivations and indicates that this likely isn’t his first encounter with child porn, as his actions were “knowing, intentional, and calculated,” as described in the lawsuit.
He was eventually arrested in March 2026, and investigators revealed that Harwood “possessed and distributed” child pornography and was charged with 8 counts of child sexual exploitation.
The maximum sentence for each count is 10 years, meaning that Harwood could face 80 years in federal prison.
While xAI does acknowledge the harm done to victims, Musk’s AI company is seeking reputational and other damages as a result of Harwood’s misuse of Grok.
xAI’s initial detection and subsequent report to law enforcement were essential evidence needed to charge Harwood, Musk’s startup claims.
The AI company claims it assisted in the arrests of at least 244 predators after its deepfake pornography scandal.
Stay updated with our latest stories and follow us on social media
Be the first to discover new stories, ideas, and updates from our team.
xAI updated its rules after Grok deepfake porn backlash
Musk’s “truth-telling” chatbot was dragged through the mud earlier this year after Grok was accused of generating deepfake pornography of women and, allegedly, children.
Grok released an apology and acknowledged an incident where it “generated and shared an AI image of two young girls (estimated ages 12-16) in sexualized attire based on a user's prompt.”
Despite the public apology, xAI and Musk did little to mitigate the damage until countries like Ireland, the UK, Malaysia, and Indonesia, as well as the European Union, investigated or banned Grok.
Musk eventually limited access to Grok and added safeguards to stop non-consensual or illegal deepfake porn requests from being fulfilled in places where it is illegal.
Following the scandal, Musk issued a stark warning to predators over X, saying “anyone using Grok to make illegal content will suffer the same consequences as if they uploaded illegal content.”
