AI disclosure requirements, platform by platform
Every platform in this category now expects you to say the account is AI. What differs is how visible that has to be, and the prominence bar sits higher than most operators assume. Below is what each one publishes, with the date we last read it.
Requirements by platform
| Platform | Synthetic personas | Disclosure required | Mechanism |
|---|---|---|---|
| Fanvue | Permitted | Yes, clear and prominent | Declared AI creator designation at signup, platform-applied AI badge, plus disclosure carried on media and in the bio. |
| OnlyFans | Not permitted | Yes, for AI-assisted content depicting the verified creator | Reporting on the 2026 updates describes an explicit label such as a tag at the start of the caption. |
| Fansly | Check current terms | Check current terms | Verify directly; we do not have a current published statement we are willing to summarise. |
| Patreon | Permitted, non-adult | Check current terms | Content restrictions matter more here than disclosure mechanics for most of this niche. |
| Your own website | N/A | Not required by Google for ranking | Google evaluates usefulness rather than production method. Disclosure is a trust decision, not an SEO requirement. |
Each cell reflects the platform's published guidance or named reporting on its 2026 updates, checked 24 August 2026.
The prominence standard
The requirement that catches people is not whether to disclose but how visibly. Fanvue's published guidance calls for clear, prominent disclosure (a watermark, caption or bio note), and its moderation is described as applying a reasonable-viewer test: would an average viewer understand what they are seeing?
That is a functional standard, not a formal one. Technically-present disclosure that a reasonable viewer would miss does not satisfy it.
- Satisfies the standard
- AI designation active on the profile so the platform badge renders. Disclosure in the first line or two of the bio. Disclosure carried on media where the media travels independently of the profile.
- Does not satisfy it
- Disclosure only in a linked terms page. Disclosure below a long bio where it is cut off. Disclosure in a font or colour chosen to be unreadable. Disclosure worded so ambiguously that it reads as a stylistic claim rather than a factual one.
- Actively breaches it
- Any representation that the persona is a real person who could be met, called or dated. This is the specific behaviour platform guidance singles out, and it is treated as misleading fans rather than as a labelling slip.
Why disclosure is in your interest
Most coverage frames disclosure as a tax on conversion. That framing misses the part that shows up on your bank statement.
- Refunds and chargebacks. A subscriber who could see the AI label before paying has a materially weaker case for a refund on the grounds that the content was not what they expected. Disclosure available pre-subscription changes what can reasonably be claimed afterwards.
- Account durability. Undisclosed synthetic accounts are a terms breach. The downside is not a warning; it is the loss of an asset with your subscriber base and balance inside it.
- Audience quality. Subscribers who knowingly pay for AI content churn differently from subscribers who feel deceived. The second group generates support load, chargebacks and public complaints.
- Optionality. A disclosed account can be promoted, discussed publicly and used as a case study. An undisclosed one cannot, because every piece of publicity is a risk.
Disclosure that works
The practical version: make it factual, make it early, make it part of the voice rather than an apology for it.
Disclosure implementation
Platform AI designation enabled and badge visible on the public profile
Check in a logged-out view. Settings showing it enabled is not the same as it rendering.
AI status stated in the first line or two of the bio
Above any fold, in the same size and colour as the rest of the bio.
Disclosure travels with media that leaves the platform
Off-platform posts do not carry your profile badge with them.
No claim, anywhere, that the persona is a real person
Including in DM scripts. This is the line that gets accounts removed.
Chatters briefed on what they may and may not say about it
Your compliance position is only as good as the least-briefed person with account access.
A dated note of what your disclosure said and when you changed it
If a dispute ever turns on what a subscriber could have seen, you will want the record.
Does a watermark count as disclosure?
Fanvue's guidance lists a watermark as one acceptable mechanism alongside a caption or bio note. A watermark has the practical advantage of travelling with the image when it is reposted elsewhere, which a bio note does not. Using both is the stronger position.
Can chatters speak as though the persona is real?
They can maintain the persona. They cannot represent it as a physical person who could be met, and the distinction is not subtle in practice. Brief this explicitly and in writing, because it is the single most likely place for your compliance position to fail.
Do I need to disclose AI use on my own website?
Google evaluates content on usefulness rather than production method, so it is not a ranking requirement. Whether to disclose is a trust decision. For a publication whose value rests on credibility, we think the answer is obvious, which is why this site's methodology page states its position.
What if the platform's own badge is enough?
It is a strong signal and it may well be sufficient on-platform. It does nothing for content that travels off-platform, which is where most of your acquisition happens. Belt and braces costs you nothing.
Where these numbers come from
Every requirement below is sourced to a platform's own published guidance or to named reporting on its 2026 policy updates, and each is linked. Enforcement practice is harder to verify than published policy and we distinguish between the two. Nothing here is legal advice.
Sources
- 01
- 02
- 03
Sacra: Fanvue company profileSource for the AI-creator revenue share quoted on this page; retrieved August 2026
- 04
Google Search: guidance on generative AI contentRelevant if you publish a website alongside the accounts
Who wrote this
AIOF Editorial DeskResearch & editorial
AIOF is an editorial desk that works from primary sources: platform terms, vendor pricing pages and published company data, read and dated rather than repeated from other write-ups. Where a figure comes from someone else we cite it and grade it. Where we have not measured something ourselves we say so, which today is most of the places you might expect a first-hand number.
- Works from primary platform documents, with the date each page was last checked published on the page itself
- Grades every figure as sourced, third-party estimate, modelled or measured, and labels which
- Publishes no first-party measured benchmarks yet, and says so on the methodology page rather than implying otherwise
- Corrections are made in place with a dated changelog entry, listed on the updates page
Revision history (2)
Re-verified Fanvue and OnlyFans requirements; added the prominence section.
First published.
Rules and pricing here change often. This page was last touched on . Found something out of date? Tell us and we will fix it in place with a dated note.
Read next
The legal lines
The content categories that carry legal exposure rather than policy risk, and why the difference matters for anyone operating AI-model accounts.
PolicyAnalysis5 minPolicyAnalysis5 minFanvue AI account setup
The full Fanvue onboarding sequence for an AI-model account: the AI declaration, identity verification, disclosure requirements and the settings that matter.
PlatformsTutorial6 minPlatformsTutorial6 minFanvue vs OnlyFans
A policy-first comparison of Fanvue and OnlyFans for AI-model accounts, including why OnlyFans is not a real option for synthetic personas.
ComparisonsComparison5 minComparisonsComparison5 min