SpaceX doesn’t make adult content. But when it filed paperwork this year to prepare for a public listing, it had to tell investors something remarkable: Grok, the chatbot built by Elon Musk’s xAI and bundled into X, carries enough legal exposure from its image-generation tools that the company has set aside $530 million to cover potential losses. The filing singled out Grok’s “Spicy” and “Unhinged” modes as a specific risk factor, warning that they “may be more irreverent and harsher” than standard offerings and could generate “exploitative, nonconsensual, or harassing content.” For a publication that covers the legitimate, subscription-based side of the creator economy, this might look like someone else’s problem. It isn’t — and the numbers explain why.
What the filing actually admits
SpaceX’s IPO paperwork acknowledged complaints that Grok’s image tools “enabled the creation and dissemination of nonconsensual explicit images and/or content representing women and/or children in sexualized contexts.” That is a corporate filing, reviewed by lawyers, choosing to disclose the risk rather than bury it — a sign of how seriously the company’s own counsel rates the exposure. The $530 million reserve was set as of December, and it sits alongside separate reporting that Apple has threatened to pull Grok from the App Store, and that French prosecutors have summoned Musk over allegations linked to child sexual abuse material and deepfake imagery generated by the platform.
The scale, in numbers
Independent researchers and court filings have put some hard figures behind the concern:
- The Center for Countering Digital Hate documented more than 23,000 apparently sexualised images of children generated in an 11-day window between late December 2025 and early January 2026 — a rate of roughly 6,000 images an hour at its peak.
- A six-month tally covering the first half of 2026 recorded 15,736 confirmed deepfake victims across 821 documented attacks, with Grok responsible for an estimated 87% of traceable synthetic attack files in that period.
- Around one in six of those attacks involved non-consensual intimate imagery of adults or children.
- xAI’s own engineers have reportedly concluded there is no reliable technical fix that allows adult-content generation while reliably blocking child sexual abuse material, because the underlying model capabilities overlap.
Regulators and courts are already moving
The legal response has been fast by the standards of tech regulation. Ofcom opened a formal UK investigation into the platform in January 2026. A Labour MP, Jess Asato, filed a High Court claim in June alleging Grok created non-consensual deepfakes. Parents in Arkansas sued in July over an alleged case of AI-generated child sexual abuse material, and a US class action that started with three Tennessee minors has since expanded to five plaintiffs across multiple states. Minnesota’s new “nudification” ban took effect on 1 August 2026, carrying penalties of up to $500,000 per incident, after a federal judge declined to block it — in response, xAI restricted Grok Imagine’s editing features for users in that state. A consultancy specialising in deepfake liability estimated total US civil exposure at over $2 billion, against roughly $7 million in currently verified direct financial losses to victims — a gap that shows how much of this is still working its way through the courts.
Why this matters if you run a legitimate OnlyFans account
None of this is really about OnlyFans. But the creator economy shares a regulatory and reputational neighbourhood with tools like Grok, and that has two practical consequences for creators who have nothing to do with any of it.
First is the direct one: creators are among the people most likely to have their photos fed into an image generator without consent. A tool capable of producing millions of sexualised images a month, with what its own maker admits is an unsolved CSAM filtering problem, is also a tool capable of generating convincing fake content of a real, working creator — undermining the subscription model that depends on subscribers paying for authentic content from a specific person. For a creator whose entire business rests on a recognisable persona, the way one creator built an entire multi-persona empire around her own likeness, an unauthorised AI copy isn’t a nuisance — it’s a direct threat to income built on trust.
Second is the spillover: payment processors, app stores and legislators tend to respond to headline-grabbing AI scandals by tightening rules across the board, not just for the platform that caused the problem. The same pattern played out when San Francisco gave Apple and Google 28 days to pull “nudify” apps — enforcement aimed at one category of bad actor that still reshapes app-store policy for everyone nearby. Legislation like the NO FAKES Act, moving through Congress partly in response to exactly this kind of story, is squarely aimed at giving creators — including OnlyFans creators — a clearer legal claim over their own likeness.
What creators can actually do
- Search periodically for your own name and likeness across image search engines and known deepfake-hosting forums, so you catch unauthorised content early.
- Keep dated, timestamped originals of your published content — they’re the clearest evidence that a circulating image is a fabrication, not something you posted.
- Use platform reporting tools immediately when you find non-consensual or AI-manipulated content, and escalate to a formal takedown request where the platform is legally required to act.
- Consider registering copyright on your original photo and video content where practical, which can strengthen a takedown or legal claim.
This article is for information only and is not legal advice. If you believe you are the victim of a non-consensual deepfake or image-based abuse, consider speaking to a solicitor who specialises in intellectual property or image rights.
