Do AI Influencers Need to Disclose They're AI? 2026 Rules

Yes, in most cases. Instagram, TikTok and YouTube require an AI label when content shows a realistic-looking person or scene, and an AI influencer built to look human falls right into that. The law depends on the country. In Brazil, no law yet requires an AI notice on a regular post, but a paid post has to be identified as advertising and can't mislead consumers, AI or not. This guide covers Brazil's rules in detail and summarizes the EU's and the US's.

This guide is informational and isn't legal advice. On October 29, 2026, we read each platform's official pages, the PDF of the guide from CONAR (Brazil's advertising self-regulation council), the bill's status in Brazil's Chamber of Deputies, and the official European Commission and FTC pages. Every rule has its source next to it, and where we didn't find an official source, we say so. Platform rules change fast: open the link before you decide.

What's mandatory and what's best practice?

Mandatory is what the platform or the law requires, with consequences if you don't comply. Best practice is what no rule requires today but that keeps you out of trouble with your audience, the brand and CONAR.

Situation	Mandatory or best practice	Source
Realistic AI video or audio on Instagram and Facebook	Label required	Meta, Feb. 6, 2024
Instagram profile featuring an AI-generated person	Must turn on the profile label, or reach may be limited	Instagram, Aug. 31, 2026
Realistic AI scene or person on TikTok	Label required	TikTok Community Guidelines, Aug. 25, 2026
Realistic AI video or Short on YouTube	Disclosure required	YouTube Help
Paid post by a human or virtual influencer, in Brazil	Must be identified as advertising	CONAR, May 26, 2026 and CDC, art. 36
Election ads with AI content, in Brazil	Notice required	TSE, via Poder360, Mar. 2, 2026
AI notice on a regular post under Brazilian law	No law in force: best practice	PL 2338/2023 in the Chamber
Bio saying the character is AI-made	Best practice	CONAR guide, note 2

All sources checked on October 29, 2026.

The ad disclosure (#ad, #publi or the paid partnership tool) and the AI label are different notices. One says the post is paid. The other says the image or video came from AI. A paid post from a realistic AI influencer needs both.

How does Instagram's AI label work?

Instagram has two labels: one on the post and, since August 2026, one on the profile.

On the post. Since February 2024, Meta has required people to use its labeling tool when they post photorealistic video or realistic-sounding audio that was digitally created or altered, and it warns that it may apply penalties to those who don't (Meta, Feb. 6, 2024). The requirement covers video and audio. For photos, labeling is best practice, and Meta applies the label itself when it detects technical AI signals in the file.

The label used to be called "Made with AI." Since July 1, 2024, it's been called "AI info." Since September 2024, it's shown prominently on AI-generated content and moves to the post's menu when AI only edited the image (Meta, updates of July 1 and Sept. 12, 2024).

On the profile. On August 31, 2026, Instagram renamed the "AI creator" label, tested earlier that year, to "AI-generated profile." It's for profiles that feature an AI-generated person instead of a human. If you only use AI tools in your creative process, you don't need it. Instagram says it will limit the reach of profiles featuring an AI-generated person that don't turn on the label, and that those who turn it on before being identified won't see a change in reach (Instagram, Aug. 31, 2026).

Flagged profiles get a notice and can check Account Status to see whether they're still eligible for recommendations. You can get back into recommendations by turning on the label or through a successful appeal. For an AI influencer, this changes the logic: the notice is no longer just post by post, it covers the whole profile.

How do you label on Instagram, step by step?

1. On the profile: tap Edit profile and turn on the AI-generated profile option. That's the path Instagram itself points to.
2. On a post or Reel: in the app, on your phone or tablet, turn on the AI label before sharing. The option isn't available on desktop, according to the Instagram Help Center.
3. If it's a paid post: turn on the paid partnership label or write #ad in the caption, visible without tapping "more."

What about Meta ads?

Ads created or heavily edited with Meta's own generative AI tools get the "AI info" label, in the three-dot menu or next to "Sponsored." When the ad has a photorealistic AI-generated human, the label sits next to "Sponsored." Meta also detects ads made with third-party AI tools through industry-standard signals (Meta, published Feb. 3, 2025, updated June 1, 2026). On top of that, Meta says that, in certain cases, advertisers themselves must disclose when they digitally create or alter a political or social issue ad (Meta).

What is TikTok's rule for AI-generated content?

TikTok's Community Guidelines, released August 25, 2026, and in effect since September 24, require creators to label AI-generated or significantly edited content that shows realistic-looking scenes or people. Unlabeled content may be removed, restricted or labeled by TikTok's team, depending on the harm it could cause. And a label doesn't make everything OK: content that's misleading about matters of public importance or harmful to individuals is still prohibited, labeled or not (TikTok Community Guidelines).

Hiding it doesn't work either. Since May 2024, TikTok has automatically labeled AI content uploaded from other platforms that carries Content Credentials metadata, from the C2PA standard (TikTok, May 9, 2024). In November 2025, it announced an invisible watermark for AI content made with its own tools and for uploads with C2PA, because the metadata can be lost when a video is edited or re-uploaded elsewhere (TechCrunch, Nov. 18, 2025).

How do you label on TikTok?

On the posting screen, tap More options and turn on the AI-generated content setting before you post (Canaltech, in Portuguese, Aug. 15, 2023). The exact name may change with the app version, but the option has been in that area since launch.

What about TikTok Shop?

A video with a TikTok Shop link is still a TikTok video, so the Community Guidelines' labeling rule applies to it too. On top of that, each country's TikTok Shop Academy can have its own rules. Brazil's, for example, has a dedicated page, Restrictions and requirements for AI-generated content (Sept. 20, 2026, in Portuguese): AI content is allowed as long as it's honest and labeled, and using AI to fabricate or exaggerate the product or to impersonate a real person is prohibited. The step-by-step for affiliate selling with an AI influencer is in AI influencer on TikTok Shop.

In the US, the rule got stricter in 2026. The US TikTok Shop Academy banned AI-generated voices, pre-recorded narration and still images in shoppable LIVEs, and began requiring at least 3 seconds of moving footage in product videos. Penalties range from commission restrictions to content removal and account bans (PYMNTS, June 8, 2026). That rule is for the US. Before going live or posting AI product videos in another country, check the TikTok Shop Academy in that country's Seller Center. The workflow for product videos with an AI influencer is on the AI videos for TikTok Shop page.

How does YouTube handle videos and Shorts made with AI?

YouTube asks you to disclose realistic content made or altered with AI, like making someone appear to give advice they didn't give or showing a realistic event that didn't happen. Beauty filters, captions and cloning your own voice for voiceovers don't need disclosure, and neither does clearly unrealistic content (YouTube Help).

Since May 27, 2026, the label appears right below the player on long-form videos and as an overlay on the video itself in Shorts. If the creator doesn't disclose and the system detects significant photorealistic AI use, YouTube applies the label itself. In some cases the label is permanent, such as content made with YouTube's own AI tools (YouTube, May 27, 2026).

Two points matter for an AI influencer channel:

• Reach and money: YouTube says the label alone doesn't change how a video is recommended or whether it's eligible to earn money (YouTube, May 27, 2026).
• Repeat offenses: creators who consistently fail to disclose may get a label applied manually and penalties, such as content removal or suspension from the YouTube Partner Program (YouTube Help).

How do you disclose on YouTube?

In YouTube Studio, on desktop or mobile, when you upload a video or Short, find the AI use setting in the Attributes section and select Yes (YouTube Help).

Does Kwai have an AI labeling rule?

As of October 29, 2026, we didn't find a public Kwai rule specifically about labeling AI-generated content. Without a built-in tool, say in the caption that the video was made with AI and keep the notice in your bio. Advertising rules don't depend on the platform: Brazil's, covered below, apply on Kwai too.

In Brazil: what does the law say about AI influencers?

Does CONAR's guide apply to virtual influencers?

Yes. The Guide to Marketing and Advertising by Digital Influencers, published by CONAR on May 26, 2026, and in effect since June 1, defines influencer advertising as content posted "on the profile of a natural person, a virtual one, animals, avatars or other characters, themes or personas, including computer-generated ones" (CONAR). An AI influencer with paid posts follows the same rules as a human one.

Three points in the guide matter if you create an AI influencer (guide PDF):

• Ad disclosure. Paid posts must be identified with the platform's paid content tool or with terms like #publicidade and #publi, visible up front without tapping "more." The guide warns that English terms like #ad may not be understood by Brazilian audiences.
• Genuine testimonials. The guide says testimonials depicting personal experiences "must be genuine." An AI character never used the product. So "I tried it and loved it" or a staged before-and-after falls under what the AI item says to avoid: simulations and testimonials that could mislead consumers about the product.
• AI notice. The guide says its rules fully apply to AI-made content, but states in a note that it doesn't create "new guidelines or duties" for that content or address notices about using the technique. Another note records that international guides stress the importance of consumers knowing they're not interacting with a real human being.

The closest thing to a precedent is the Volkswagen commercial featuring singer Elis Regina recreated with AI. In August 2023, CONAR's Ethics Council dismissed the complaint. On the lack of an AI notice, it decided 13 to 7 that "transparency is a fundamental ethical principle" and that it had been respected in that case, because the use of the tool was "evident in the ad" (Poder360, Aug. 24, 2023). The practical takeaway: an AI influencer made to look real doesn't make the AI use evident, so the notice matters more.

What does Brazil's Consumer Protection Code say?

Brazil's Consumer Protection Code (CDC) applies to advertising on any platform and with any technology (Law No. 8,078/1990):

• Art. 36: advertising must be identifiable by consumers "easily and immediately."
• Art. 37: bans misleading advertising, including by omitting essential product information (§ 3).
• Art. 38: whoever sponsors the ad has to prove it's truthful.
• Arts. 56 and 67: violations can bring administrative fines, and running advertising known to be misleading is a crime, punishable by three months to one year of detention and a fine.

In practice, the bigger risk isn't AI itself. It's what AI makes look true: a result the product doesn't deliver, a usage testimonial that never happened, a staged demo.

Is Brazil's AI bill (PL 2338/2023) law yet?

No. The Senate passed the bill on December 10, 2024 (Senate), and it moved to the Chamber of Deputies. The rapporteur in the special committee is Deputy Aguinaldo Ribeiro (PP-PB), appointed on May 20, 2025. On October 29, 2026, the Chamber's page showed the status "Aguardando Parecer do(a) Relator(a) na Comissão Especial" (awaiting the rapporteur's report in the special committee), and the latest actions, on October 26 and 27, were attachments of other bills (Chamber of Deputies). If the Chamber changes the text, it goes back to the Senate before it can be signed into law.

What about the TSE's election rule?

In election advertising, an AI notice is mandatory. Resolution No. 23,755, approved by the TSE (Brazil's Superior Electoral Court) on March 2, 2026, updated the advertising rules for the 2026 election. Content made or altered with AI needs an explicit notice. From 72 hours before until 24 hours after the vote, new AI content that changes the voice or image of a candidate or public figure can't be circulated, even if it's labeled. The rule also bans AI systems from recommending candidates (Poder360, Mar. 2, 2026; Estadão, via Money Times, Mar. 3, 2026).

For an AI influencer, the practical rule is to keep the character away from candidates, parties and elections. Brazil's next election is the 2028 municipal one, and the rules are revised before each election.

Can I use a face that looks like a real person?

This is where AI influencers get the most complicated. Brazil's Civil Code lets a person forbid the use of their image for commercial purposes, without prejudice to damages (art. 20), and bans using someone's name in advertising without permission (art. 18) (Civil Code). Súmula 403 of the STJ (Brazil's Superior Court of Justice) says compensation for unauthorized commercial use of someone's image doesn't depend on proof of harm (STJ).

In practice: don't generate a face "inspired" by a celebrity or a well-known person, and don't use another person's photo as a base without written permission. If you want to start from your own face, Avatrix's character creator accepts a photo of you or of someone who authorized its use.

Outside Brazil: what do the EU and the US require?

European Union. Article 50 of the AI Act (Regulation (EU) 2024/1689) applies from August 2, 2026. Anyone using AI professionally, including a person who regularly earns money from it, must disclose deepfakes no later than first exposure, with a clear label people can see or hear; the machine-readable mark from the AI tool isn't enough on its own. A deepfake is AI image, audio or video that resembles people, places or events that exist or could plausibly exist and that can pass as authentic, so a photorealistic AI influencer can fall under the definition. For evidently artistic or fictional work, an appropriate notice that doesn't hamper the work is enough (European Commission).

United States. On June 29, 2023, the FTC updated its Endorsement Guides to clarify how they apply to fake reviews and virtual influencers, and warned that a platform's built-in disclosure tool might not be adequate (FTC). Its Consumer Reviews and Testimonials Rule, in effect since October 21, 2024, doesn't ban AI avatars, but it does ban fake testimonials, like lying about having used a product, and using a celebrity's avatar without permission to praise a product as if the celebrity endorsed it (FTC).

What happens if you don't label?

Where	What can happen	Source
Instagram, on posts	Meta may label the content and apply penalties	Meta, Feb. 6, 2024
Instagram, on profiles	An unlabeled profile featuring an AI-generated person may have its reach limited and drop out of recommendations	Instagram, Aug. 31, 2026
TikTok	Removal, restriction or a label applied by TikTok's team	TikTok Community Guidelines, Aug. 25, 2026
YouTube	Automatic label; creators who repeatedly fail to disclose may have content removed or be suspended from the Partner Program	YouTube Help
TikTok Shop, US	Commission restrictions, content removal and account bans	PYMNTS, June 8, 2026
CONAR (Brazil)	Recommendation to change or pull the ad	CONAR guide, 2026
Consumer Protection Code (Brazil)	Administrative fine and, for advertising known to be misleading, a crime	Law No. 8,078/1990

Sources checked on October 29, 2026.

What's the checklist before posting with an AI influencer?

1. Instagram profile with the AI-generated profile label turned on.
2. Bio saying, in plain language your audience understands, that she's an AI-created character.
3. AI label on every realistic post and video: the AI label on Instagram, AI-generated content on TikTok and AI use on YouTube.
4. Paid posts with the paid partnership label or #ad, visible without tapping "more."
5. No usage testimonials the character couldn't have had, and no staged before-and-afters.
6. An original face: don't copy a celebrity or a real person. Someone else's photo only with written permission.
7. No politics: no candidates, parties or elections coming out of the character's mouth.
8. TikTok Shop: check your country's Academy rules before going live or posting AI product videos.
9. Records: keep the brand contract and a screenshot of each paid post with the disclosure. CONAR's guide says it's relevant to keep records of the measures taken.

How do you start your AI influencer the right way?

Start with the character and finish with the label. The step-by-step for creating the influencer, with the same face in every photo and video, is in the guide to creating an AI influencer without showing your face, which includes a summary of these rules. To see how Brazilian brands built theirs, read the Brazilian virtual influencer case studies. On Avatrix, you create the face, photos and videos of your AI influencer. When it's time to post, the label is up to you.

Create my AI influencer →