An AI disclosure statement is a short line that tells your audience the content was created or altered with AI. That is the whole idea. The hard part is not writing it, it is knowing where it is required, what it has to say, and where to put it so it actually counts. Below are copy-paste templates for a profile, a post, a video and an About page, followed by what the FTC, the platforms and the EU AI Act actually ask for.
What is an AI disclosure statement?
It is a plain notice that AI was involved in making what someone is about to look at. It does two jobs at once. It keeps you on the right side of platform rules and advertising law, and it builds trust with an audience that increasingly assumes the worst about unlabeled synthetic content. For a persona that is openly AI, the disclosure is part of the brand. You are the account that labels plainly while half the field pretends otherwise.
A good disclosure is specific and easy to see, placed where the viewer meets it early. "Made with AI" at the top of a bio beats a paragraph of legalese no one reads, and both beat a label hidden where the viewer will never scroll.
AI disclosure statement examples (copy-paste templates)
Use these as starting points and adjust the wording to your voice. Keep them where a viewer meets them early.
Profile or bio (account-level):
Virtual character. All images on this profile are created with AI.
Short post caption:
Created with AI. This is a digital character, not a real person.
Detailed post caption (with a brand connection):
AI-generated image. This is a virtual character. Paid partnership with [brand].
Video description (YouTube and similar):
This video contains realistic content that was altered or generated with AI. The character shown is a digital persona, not a real individual.
About page (website or media kit):
[Persona name] is a virtual character. Every image and video is generated with AI. There is no real person behind the photos. Where a post involves a paid or sponsored relationship, that is disclosed in the post itself.
Newsletter or email footer:
Imagery in this email is AI-generated. [Persona name] is a digital character.
Where are you legally required to disclose AI content?
Two layers matter for most creators: advertising law and the EU AI Act.
On advertising, the FTC treats synthetic personas the same as human ones. Its Endorsement Guides reach virtual influencers, fake reviewers and anything that presents as a person, so if a brand pays for or controls an AI-generated endorsement, the material connection has to be disclosed exactly as it would for a human creator. The FTC also treats AI-generated testimonials as fake reviews, since a testimonial is supposed to reflect a real person's honest experience. Under its Penalty Offenses authority the FTC can seek civil penalties of up to $53,088 per violation (the 2025 ceiling, adjusted annually for inflation), and each non-compliant post can count separately.
On the EU side, the AI Act's transparency obligations under Article 50 apply from 2 August 2026. They require that AI-generated or manipulated image, audio and video content, including deepfakes, be disclosed as artificially generated, with a limited carve-out for evidently artistic or satirical work where the disclosure must not spoil the piece. If any of your audience is in the EU, this is the rule that will shape your labeling, covered in full in the EU AI Act deepfake rules.
(This is general information, not legal advice. Rules differ by country and several US states have added their own AI disclosure laws; check the current position for your markets.)
What do the platforms require?
Platform rules move faster than the law and are enforced more often. The current shape:
| Platform | What needs a label | How | If you skip it | |---|---|---|---| | YouTube | Realistic altered or synthetic content (real-looking people, places, events). Production-assistance use is exempt. | "Altered or synthetic content" disclosure in Studio; adds a description label | Label applied for you; penalties or Partner Program removal for repeat cases | | TikTok | Realistic AI-generated content (people, scenes, audio) | Creator toggle, plus automatic labeling from C2PA Content Credentials and detection | Strictest of the three: violations can lead to removal and account strikes, with less warning-first tolerance | | Instagram / Meta | AI-generated or modified content | Automatic "AI info" label from metadata, plus an opt-in "AI creator" account label (rolling out in 2026) | Automatic detection can label content; detection is acknowledged as imperfect |
The pattern across all three is the same: realism is the trigger, the platform would rather you self-label, and several of them now read provenance metadata and will label you whether or not you disclosed. Labeling yourself first is the only way to control how it reads. For the exact per-platform rules, see disclosing AI content on Instagram and YouTube.
How do you write a disclosure that actually counts?
Four rules cover most of it:
- Make it clear and conspicuous. Plain language, visible without effort, not hidden in a hashtag wall or a link no one opens.
- Put it upfront. Profile bio for the account, top of the caption for a post, the description and native label for a video.
- Say what is AI. "Created with AI" or "AI-generated image" beats a vague "edited." If the persona is fully synthetic, say there is no real person behind it.
- Disclose money separately. A content label is not an ad disclosure. If there is a paid or material connection, call it out on its own, in the post.
What this looks like for an AI influencer or virtual persona
For a one-off post, disclosure is a line. For a persona you run continuously, it is a setting you configure once and let define the account. Put the account-level statement in the bio, keep a standard caption label ready, turn on each platform's native AI label, and add the paid-partnership disclosure on the posts that need it. Done this way, disclosure stops being a chore per post and becomes part of what the persona is.
That is the approach Cladegrove is built around: characters that are openly AI, run consistently, with the labeling baked into how you operate them rather than bolted on after. If you are setting one up from scratch, how to create an AI influencer walks the build, and the companion piece on who owns the copyright of AI images covers the ownership side of the same question.
See how Cladegrove handles disclosed, consistent personas on the AI disclosure page, or start with running an AI influencer.
Disclosure used to feel like the boring compliance corner of this work. It is not. The accounts that say plainly what they are tend to get more trust, not less, because the audience already suspects and resents being played. Labeling first, every time, turned out to be a feature of the persona rather than a tax on it. None of this is legal advice; for your own situation, talk to a lawyer.
Fabio Ariotti, operator
Common questions
Do I have to disclose AI on every post?
Not always. The trigger most platforms use is realism: realistic AI content that a viewer could mistake for a real person or event needs a label, while clearly unreal or AI-assisted production work usually does not. The safer habit for an openly AI persona is to disclose at the account level once and label individual posts where realism could mislead.
Where should the disclosure go?
Upfront and visible, not buried. Put it where a viewer sees it before or as they consume the content: in the profile bio, at the top of a caption, in the video description and the platform’s own AI label. A disclosure hidden at the end of a hashtag block does not count as clear and conspicuous.
Is "made with AI" enough?
As a content label, often yes. As an advertising disclosure, not by itself. If the post is an endorsement or has a paid or material connection behind it, regulators expect you to disclose that relationship too, separately from the fact that the image is AI-generated.
What happens if I do not disclose?
It depends on who is enforcing. Platforms can apply a label for you, demote, demonetize, or strike repeat offenders. Advertising regulators can treat undisclosed paid endorsements as deceptive, with civil penalties. In the EU, the AI Act adds its own transparency duties for synthetic and deepfake content from August 2026.
