Grok Scandal Deepens — Who’s Really Liable?

Empty courtroom with wooden furnishings and judges bench.

Elon Musk’s xAI is suing a Grok user for allegedly creating child sex abuse images with its AI tool at the very same time the company faces mounting accusations that Grok itself made thousands of sexualized images of minors.

Story Snapshot

xAI says South Carolina man Terry Wayne Harwood used Grok to turn normal photos of kids and adults into explicit deepfakes, breaking its rules.
Harwood was already arrested earlier this year on multiple felony charges for sexual exploitation of minors, giving prosecutors a criminal backdrop.
xAI faces separate lawsuits and city and teen complaints saying Grok’s “spicy mode” helped create tens of thousands of sexualized images of minors.
New “platform versus user” lawsuits over AI deepfakes fuel a deeper fear shared by both left and right: powerful tech and government systems are failing to protect children and basic privacy.

xAI’s lawsuit against a Grok user

Elon Musk’s artificial intelligence company xAI has filed a 12-page lawsuit in federal court in Texas against Terry Wayne Harwood, a man from South Carolina. The complaint says Harwood opened at least two Grok accounts using false identities and then used the chatbot to turn non-sexual photos of adults and minors into explicit sexual images without their consent. Court documents claim he designed “misleading prompts” to get around Grok’s safety tools that are supposed to block child sexual abuse material.

xAI’s filing says Harwood uploaded regular photos of several people, including a fully dressed girl who looked about 10 or 11 years old, and asked Grok to sexualize them. The company accuses him of generating, possessing, and sharing child sexual abuse material and non-consensual intimate imagery of both minors and adults using Grok. xAI argues that this behavior broke its terms of service and user agreement, and it is asking the court for money damages and a permanent order blocking Harwood from using Grok in the future.

Criminal charges and what xAI claims it did

Separate from the civil case, Harwood was arrested in South Carolina on February 26, 2026, and charged with eight felonies for sexual exploitation of minors. Reports say the charges include several counts of second- and third-degree sexual exploitation, which involve creating, possessing, or sharing illegal sexual images of children. xAI says its internal monitoring and detection efforts helped lead to Harwood’s arrest, suggesting the company was tracking suspicious Grok activity and reported what it found to authorities.

Even so, parts of the complaint show limits in what has been proven so far. For key claims about child sexual abuse material, xAI uses the phrase “upon information and belief,” which usually means the company is relying on outside reports or inference rather than presenting full forensic proof of each image in the public filing. The lawsuit also does not say how much money xAI is seeking in damages, and it does not include the exact prompt text or detailed technical description of how Grok’s safeguards were bypassed.

Grok’s “spicy mode” and lawsuits against xAI

While xAI sues Harwood for abusing Grok, the company itself is already under heavy legal fire for what critics say are its own unsafe design choices. Teenagers in Tennessee filed a class-action lawsuit in March, saying Grok’s image tools were used to turn real photos of them, taken when they were minors, into explicit deepfake images. Their complaint argues xAI designed Grok to create sexual content for profit and failed to build strong protections to stop the tool from making sexual images of children.

Another complaint from Baltimore’s mayor and city council says Grok flooded the X platform with non-consensual intimate images and child sexual abuse material, putting city residents at risk whenever photos of them or their kids are uploaded. That filing cites research from the Center for Countering Digital Hate, which estimated Grok produced around 23,000 sexualized images of minors in just 11 days during a period when xAI promoted a “spicy mode” for more sexual content. Together, these suits argue xAI did not just miss a few bad actors, but knowingly released a powerful tool that could undress and sexualize real people, including kids, with minimal friction.

Elon Musk’s public denial and growing global backlash

Elon Musk has publicly tried to push back on the idea that Grok is a factory for child abuse material. In January, he claimed on X that Grok had generated “literally zero” naked images of underage people. Critics now point to that statement as conflicting with later complaints and outside research that describe large numbers of sexualized images of minors created with Grok, especially during the “spicy mode” period. This gap between Musk’s words and what lawsuits allege feeds a sense that powerful tech leaders are not being straight with the public.

Pressure on xAI and X is not just coming from private lawsuits. Media and legal reports say countries like Malaysia and Indonesia have banned Grok, and the United Kingdom’s regulator Ofcom has opened a criminal investigation into whether X broke British law by allowing non-consensual sexualized deepfakes. In the United States, dozens of state attorneys general have raised alarms about AI tools that can erase clothing, sexualize minors, and spread fake intimate images. Many ordinary Americans see these moves and wonder why it took foreign governments and scattered lawsuits to force basic guardrails, instead of their own federal system stepping in early.

What this case says about power, platforms, and the “deep state” fear

Legal experts say the Harwood lawsuit is part of a new trend: AI platforms suing individual users to show they are not the main offenders when illegal content appears. The strategy leans on contract law, arguing that users broke platform rules, and aims to create court precedents that protect companies from civil blame even as their tools can be abused. That may help xAI in court, but for many people on both the right and the left, it looks like another example of big players shifting responsibility downward while keeping profits and control.

Context the headlines skipped: three Tennessee minors sued xAI on Monday, alleging Grok generated CSAM from their photos. xAI filed against this user the next day. Draw your own conclusions about which lawsuit caused which.

— Corey Quinn (@QuinnyPig) July 16, 2026

Conservatives angry about “woke” tech and liberals worried about corporate power share a basic worry here: the systems meant to protect kids and basic privacy feel rigged and slow. When an AI tool can undress a 10-year-old in a photo, or turn a teen’s selfie into porn, families expect strong federal action, not finger-pointing between companies and lone users. The Harwood case and the broader Grok scandals highlight how far law and policy lag behind fast-moving technology, and feed the belief that elites in both industry and government are more focused on covering themselves than on truly safeguarding the public.