Key Takeaways
- AI influencers are legal in the US when you disclose paid partnerships under FTC rules, respect right-of-publicity limits, accept that fully AI-generated content is not copyrightable, and follow each platform’s synthetic-content labeling policy.
- The FTC expects two disclosures on every paid post: one for the commercial relationship (“ad,” “sponsored”) and one stating the influencer is AI-generated. Both must be clear and conspicuous.
- New York’s AI Synthetic Performers Disclosure Law, effective June 9, 2026, requires conspicuous AI labeling in ads reaching New York consumers and authorizes fines up to $5,000 per violation.
- TikTok, Instagram, and YouTube each require visible AI labeling at upload or profile level. Missing labels can reduce reach, trigger content removal, or lead to account penalties.
- Sozee helps creators apply these rules in practice by generating original characters, keeping likeness consistent, and maintaining a documented asset trail that supports human authorship claims.
Create A Compliant AI Influencer With Sozee
The FTC Rules For AI Influencers: Two Disclosures Per Post
The FTC’s Endorsement Guides (16 CFR Part 255), revised effective July 26, 2023, and the plain-language Disclosures 101 for Social Media Influencers together govern every AI influencer account operating in the US. The standard is that disclosures must be “difficult to miss and easy to understand,” and endorsements must not be deceptive.
The 2023 revision explicitly expanded “endorsement” and “endorser” to cover AI-generated content and computer-generated virtual influencers. A synthetic persona now faces the same standard as a human creator.
Many creators miss compliance because they merge two separate obligations. Disclosing a paid partnership and disclosing that the influencer is AI are distinct duties. A post can satisfy one and violate the other. Both must appear clearly and conspicuously in the same place as the content, not buried in a bio, an end-of-caption hashtag block, or behind a “more” link.
Plain words such as “ad,” “advertisement,” or “sponsored” satisfy the disclosure standard, while vague shorthand like “sp” or “collab” fails it. The disclosure duty falls on the operator, because a fictional persona cannot hold legal responsibility, so any enforcement action names the business behind the account.
The FTC’s Consumer Review and Testimonials Rule (16 C.F.R. Part 465), effective October 21, 2024, carries civil penalties of up to $53,088 per violation for creating, selling, procuring, or disseminating materially false consumer reviews or testimonials. In December 2025, the FTC issued formal warning letters to nearly a dozen companies, marking the agency’s first public use of the Rule as an enforcement tool.
What This Means For You: Label every paid post with plain words like “ad” or “sponsored” in the first line, and avoid personal testimonials that an AI influencer cannot truthfully give.
Build A FTC-Compliant AI Influencer With Sozee
New York’s June 9, 2026 AI Disclosure Law: Concrete Deadlines And Fines
New York’s statute is the first in the US to attach a fixed start date and a per-violation fine to AI ad disclosure. The table below highlights the provisions that matter most if you run paid campaigns that can reach New York consumers.
| Detail | Provision |
|---|---|
| Law | New York AI Synthetic Performers Disclosure Law, N.Y. Gen. Bus. Law § 396-b |
| Signed | December 11, 2025, as Chapter 617 of the Laws of 2025 |
| Effective date | June 9, 2026 |
| Penalty | $1,000 for a first violation; $5,000 for each subsequent violation |
| Who it applies to | Any person engaged in the business of dealing in property or a service who produces or creates an advertisement for a commercial purpose, with actual knowledge that a synthetic performer appears, if the ad reaches New York consumers, regardless of where the advertiser is headquartered |
| What must be disclosed | A conspicuous disclosure in the advertisement itself that a synthetic performer appears |
| Key carve-outs | Audio advertisements (subdivision 7(a)); expressive works (subdivision 4); media carriers (subdivision 8) |
A “synthetic performer” is a digitally created asset made with generative AI or a software algorithm. It must create the impression of a human performance and must not be recognizable as any identifiable natural performer. Digital replicas of real people fall outside this statute and are governed instead by New York’s Civil Rights Law, including § 50-f.
New York is not acting alone. If your ads reach audiences outside the US, the same disclosure logic applies with different deadlines and penalties. The EU AI Act Article 50 transparency obligations apply from August 2, 2026, with fines up to €15 million or 3% of worldwide annual turnover. India’s amended IT Rules took effect February 20, 2026 and cover audio, which New York does not.
What This Means For You: If you advertise to New York audiences, disclosure is mandatory and the fine range is significant. Audit every live creative before it runs.
Does TikTok Allow AI Influencers? Platform-By-Platform Synthetic-Content Policies
TikTok, Instagram, and YouTube all allow AI influencers, yet each handles synthetic-content labeling differently. The summaries below show what you must do on each platform.
TikTok
TikTok requires visible on-video labeling of AI-generated visuals and audio depicting realistic people or scenes and exempts AI-assisted text such as scripts, captions, and hashtags. It auto-detects AI via C2PA Content Credentials and invisible watermarking. Unlabeled AI “may be auto-flagged, suppressed, or removed,” and repeated violations can affect account standing.
On August 31, 2026, Instagram renamed its “AI creator” label to “AI-generated profile” and stated it will limit the reach of profiles featuring an AI-generated person that do not add the label. Creators who proactively add the label will not see a change to their profile’s reach, while creators who do not appropriately label their AI-generated profiles may see limits to their profile’s reach.
The label is a profile-level toggle, distinct from the post-level “AI info” label. Using AI to edit photos or polish captions does not require the profile-level label.
YouTube
YouTube requires creators to disclose realistic altered or synthetic media via the Creator Studio tool. Creators who consistently fail to disclose may face manual labeling, content removal, or suspension from the YouTube Partner Program. YouTube’s model is deception-based rather than a blanket ban, and its separate inauthentic-content policy makes mass-produced, unoriginal AI content ineligible for monetization regardless of disclosure.
Exempt from disclosure: beauty filters, color and lighting filters, special-effects filters, caption creation, clearly unrealistic content such as animation, and cloning your own voice for voiceovers.
What This Means For You: Add AI labels at upload using each platform’s native tools and confirm policy details on the platform’s help pages before you schedule.
Is It Illegal To Create AI Videos Of Celebrities? Right-Of-Publicity Limits
AI videos of celebrities without consent create one of the fastest paths to a civil lawsuit and, in some states, to criminal exposure. Roughly 25 states have specific right-of-publicity statutes and about 38 offer some form of protection through statute or common law. California Civil Code § 3344 covers name, voice, signature, photograph, or likeness used for advertising without prior consent. New York’s protection is exclusively statutory under Civil Rights Law §§ 50 and 51.
The foundational precedent is White v. Samsung Electronics America, 971 F.2d 1395 (9th Cir. 1992). A robot in a blond wig beside a game-show set was held to have misappropriated Vanna White’s persona even though the ad never used her name or photograph. Identifiability does not require a literal photograph, because context and evocation can be enough.
On the question of whether you can be sued for using AI-generated images, the risk appears when the image is readily identifiable as a real person and is used for a commercial purpose without consent. An original AI-generated character with no source photos carries no such exposure because there is no identifiable natural person to appropriate.
What This Means For You: Build an original character or get written consent, and avoid training on a real person’s likeness without permission.
Copyright And Human Authorship: What You Actually Own
The U.S. Copyright Office’s March 16, 2023 Registration Guidance for Works Containing AI-Generated Materials (88 Fed. Reg. 16,190) and Part 2 of its Report on Copyright and Artificial Intelligence (January 29, 2025) set the standard. Works generated wholly by AI cannot be registered. AI-assisted works are evaluated case by case, and protection extends only to the human-authored portions.
Thaler v. Perlmutter, 130 F.4th 1039 (D.C. Cir. 2025) held the Copyright Act “requires all eligible works to be authored in the first instance by a human being.” The Supreme Court denied certiorari in March 2026, which closed the door on AI authorship claims.
Creators can protect original character design decisions, scripts, edits, and creative selection, coordination, and arrangement. The Office registered “A Single Piece of American Cheese” while expressly excluding its AI-generated image components, covering only the human arrangement.
What This Means For You: Keep records of your creative direction, prompts, and edits, because documented human authorship is the portion you can protect.
How To Tell If An Influencer Is AI: Disclosure In Practice
From the creator’s side, the label belongs where viewers see it before they consume the content. Place it on the profile, on the video, or in the first line of the caption, not buried in a bio or behind a “more” link.
The plain-language rule from the FTC section applies here too: use “ad” or “sponsored,” not “sp” or “collab.” For the persona itself, a clear statement that the person featured is AI-generated is required wherever platform policy or law mandates it.
Making disclosure part of the brand rather than a disclaimer builds audience trust. A labeled AI creator who is open about the format avoids the backlash that follows undisclosed profiles. Instagram’s August 2026 policy change responded directly to user frustration at discovering profiles were AI-generated without any warning.
What This Means For You: Put the label where viewers see it first and treat disclosure as part of your brand voice.
Sozee: The Production Platform That Makes Compliance Practical
The rules above share one practical requirement: you need an original character and a documented asset trail that proves human direction. That is the gap Sozee was built to close.
Sozee generates an entirely original character from scratch, which eliminates right-of-publicity exposure because there is no identifiable natural person to appropriate. That originality only holds up if the character stays consistent, so Sozee locks likeness across frames and sets. Consistency then produces a documented trail, because reusable environments, outfits, and objects are stored in the Vault alongside every image and video. This structure gives creators the evidence of human creative direction that the Copyright Office’s authorship standard rewards.
General-purpose AI tools often produce a different face every time and leave no consistent asset trail, which makes brand consistency and authorship documentation difficult. Sozee builds compliance and verification into setup. Its native scheduling and analytics also separate what Sozee posted from what the creator posted, producing a clean record for any audit.
| Approach | Right-Of-Publicity Exposure | Documented Authorship Trail |
|---|---|---|
| Original AI character built in Sozee | None, because no identifiable natural person is depicted | Yes, because Vault retains every asset, environment, and outfit as a reusable record supporting human creative direction |
| AI likeness trained on a real person without consent | High, and actionable under state right-of-publicity statutes with statutory damages from $750 | No, because output is unprotectable and the source likeness is unauthorized |
| General-purpose AI tool with no locked likeness | Varies, because no consistent character means no consistent exposure analysis | No, because there is no reusable asset trail and no evidence of human creative direction |
Produce And Track AI Influencer Assets In Sozee
Pre-Launch Compliance Checklist
- Confirm your character is original or consented, and avoid using a real person’s likeness without written permission. See the full workflow in How To Create An AI Influencer: The 8-Step Playbook and How To Protect Privacy When Creating AI Virtual Influencers.
- Label every paid partnership with plain words in the first line, using “ad,” “advertisement,” or “sponsored.”
- Label AI generation per platform at upload and avoid relying on caption text alone.
- Check New York and international disclosure triggers before any ad reaches those audiences.
- Document human authorship by saving prompts, drafts, edits, source files, and version history. Review Commercial Rights For AI Influencer Photos & Videos for what you can and cannot own.
- Keep records of every asset in a single, retrievable location. Sozee’s Vault is built for this, and How To Create Consistent AI Virtual Influencer Content walks through the full asset-management workflow.
Frequently Asked Questions About AI Influencer Compliance
Can I Be Sued For Using AI-Generated Images?
Yes, if the image is readily identifiable as a real person and is used for a commercial purpose without consent. As noted above, statutory damages start around $750 in many states, and the identifiability threshold can be lower than many creators expect. An original AI-generated character with no source photos avoids that exposure.
Does TikTok Allow AI Influencers?
Yes. TikTok allows AI influencers but requires visible on-video labeling of realistic AI-generated visuals and audio. As covered above, TikTok auto-labels via C2PA Content Credentials and has labeled billions of videos. The practical takeaway is to self-label at upload and treat auto-detection as a backstop.
Is It Illegal To Create AI Videos Of Celebrities?
AI celebrity videos are usually actionable civilly and can trigger criminal exposure in some states. New Hampshire’s Fraudulent Use of Deepfakes law makes it a Class B felony to knowingly create or distribute a deepfake for the purpose of harassing, defaming, or causing financial harm, and Washington prohibits the knowing distribution of a forged digital likeness with intent to defraud or harass. The federal NO FAKES Act advanced out of the Senate Judiciary Committee on June 18, 2026 and would create a federal property right in an individual’s voice and visual likeness if enacted. The safest path remains an original character with no source material drawn from a real person.
How Do I Tell If An Influencer Is AI?
Look for a profile-level label such as Instagram’s “AI-generated profile,” an on-video disclosure on TikTok, or a Creator Studio disclosure on YouTube. If none appears and the person is photorealistic, treat the account as undisclosed and proceed cautiously. From the creator’s side, the label must appear where it is seen before the content is consumed, and vague shorthand like “sp” or “collab” fails the FTC’s clear-and-conspicuous standard regardless of platform.
Conclusion: Disclosure Requirements Keep Expanding
New York’s law took effect June 9, 2026. The EU’s Article 50 obligations apply from August 2, 2026. The NO FAKES Act advanced out of the Senate Judiciary Committee on June 18, 2026. Each development points toward broader disclosure and tighter enforcement. Creators who build compliant, original, locked-likeness AI influencers with a platform like Sozee will be the ones who scale, because they will not be rebuilding after a takedown, a fine, or a lawsuit. For the complete production workflow, see How To Create An AI Influencer In 2026: Complete Guide.
Launch Your AI Influencer Safely With Sozee