What the filing records, and what it asserts
A family filed the second federal suit against xAI on Thursday in the Fayetteville division of the U.S. District Court for the Western District of Arkansas, on behalf of a 16-year-old boy. According to the complaint, the images that depict him were generated with Grok, the artificial intelligence model xAI builds, from authentic photographs the Bentonville photographer Russell Bloodworth had taken of the child, and were then disseminated widely online. Bloodworth faces 100 counts of pandering or possessing visual material depicting sexually explicit conduct involving a child and 100 counts of unlawful creation or distribution of deepfake material; he has pleaded not guilty and is awaiting trial.[1]
None of that is yet a finding. The criminal counts are charges, the civil filing is a set of allegations, and the sentence the complaint places at its centre — that this was a corporate choice to monetize a system xAI could reasonably have foreseen would be used this way, and that the company had been put on notice earlier and did not act — is the part that needs a document behind it. At least four related cases now carry claims that mirror one another, in Benton County Circuit Court, in Fayetteville and in Little Rock. Repetition across dockets tells you how many families have sued; it adds nothing to the evidence for the notice allegation.[1]
Four requests, four points on the chain
The complaint asks for four things, and each one lands somewhere different. Compensatory, actual and punitive damages answer the harm already done. Restitution and disgorgement of all revenue and profits attributable to the unauthorized use of the child's likeness answer the money. A permanent injunction covers the copies. And a separate order would require xAI to implement reasonable technical safeguards sufficient to prevent the generation of nonconsensual, sexually explicit depictions of identifiable real people, including minors.[1]
The injunction is where the two standards separate. For copies xAI holds, the complaint asks for removal and permanent deletion — a result the company can be measured against. For material already spread to other platforms, it asks for commercially reasonable efforts to take it down, which measures conduct rather than outcome, and the difference matters most for exactly the copies that have already left. Only the safeguard order reaches back to the step where an identifiable person's likeness enters the model. A judge could still read the effort standard strictly enough to require an auditable takedown log, which would narrow the gap; the parties could also settle on a removal process that never becomes a public document at all.[1]
Which document would settle the notice claim?
An earlier column in this space read two motions to dismiss in the publishers' case against Anthropic and found that neither side was arguing over a finding; both were arguing over which document the complaint had to show, and different claims demanded different objects. The same test applies here, and it points somewhere narrow. Damages need harm, disgorgement needs revenue attribution, the injunction needs possession. The notice claim needs something else entirely: a dated record of what xAI was told about Grok being used this way, and when.[1], [2]
That document is checkable rather than rhetorical. When xAI responds to the Fayetteville complaint, its filing will either dispute prior notice or concede it, and whichever way it goes, the file will show whether a dated notice record is attached to the allegation or whether the allegation still stands on the plaintiffs' description of it. Until then, the strongest thing that can be said about the four mirrored complaints is that they agree with each other.[1]