OnlyFans rewrote its rules on AI-generated content in 2026, and the changes are significant enough that “I didn’t realise that counted as AI content” is now a real risk to a creator’s account, not just a theoretical one. Here’s exactly what changed and what it actually requires of you.
The headline change: zero tolerance for deepfakes
Any AI-generated media that fabricates or replaces a real person’s likeness without their consent is now banned outright – full stop, regardless of who the target is. That covers deepfakes of public figures, of other creators, and of private individuals equally. Violations carry permanent account removal, and the policy explicitly allows for legal referral in serious cases.
Fully fictional AI characters – synthetic personas with no real-person basis – also can’t be monetised on the platform. This closes a grey area some creators had been operating in: an entirely AI-generated “creator” is off the table as a business model here, not just non-consensual use of a real person’s likeness.
What’s actually still allowed
Creators can use AI-generated content built on their own, verified likeness. The condition is disclosure: a visible “#ai” or “#AIGenerated” tag has to appear where a subscriber sees it before paying – not tucked into a bio, a pinned post, or a highlight where it’s easy to miss. Collaborator content requires written consent, filed with the platform in advance of posting, if another verified creator’s likeness is involved.
What happens if you get it wrong
Undisclosed AI content – even content using your own likeness, if the disclosure isn’t visible at the point of purchase – triggers strikes, revenue holds, or demonetisation. Repeated violations escalate to suspension. There’s also a US-specific compliance layer worth knowing about: 18 U.S.C. §2257 record-keeping requirements apply to AI content depicting real people, meaning seven years of documentation retention, with federal criminal penalties attached to non-compliance for platforms and, in some circumstances, creators.
One detail that trips people up: the platform’s rules explicitly note that fully automated chatbots impersonating a creator without human review also violate the policy. If you’re using an AI chat tool to handle fan messages, “set it and forget it” isn’t compliant – human oversight is a stated requirement, not a nice-to-have.
Why this happened now
This overhaul tracks a broader, platform-wide reckoning with deepfakes across the creator economy in 2026 – the same pressure that’s driven detection and takedown tools to mature rapidly this year. OnlyFans tightening its own rules is as much about limiting the platform’s legal exposure as it is about protecting individual creators, but the practical effect for working creators is the same either way: clearer lines, and less ambiguity about what’s compliant.
What to actually do
- If you use AI tools on your own likeness, tag every piece visibly and before the paywall – not as an afterthought.
- Get any collaborator consent in writing and filed before you post, not after.
- If you use any AI chat or auto-response tool for fan messages, make sure a real human is genuinely reviewing what it sends, not just spot-checking occasionally.
- If someone else is using AI to fake your likeness, that’s the flip side of this policy working in your favour – see our guide on protecting yourself from deepfakes and AI impersonation for the practical steps.
