How to Disclose AI-Generated Ads on Meta and TikTok (2026)

Why Disclosure Requirements Matter for AI Ad Creators

AI-generated video ads — avatars, synthetic voices, AI-written scripts overlaid on AI-generated visuals — are now mainstream in performance marketing. Tools like Arcads, Creatify, and MakeUGC have made it possible for small teams to produce hundreds of ad variants per week. But the same capabilities that make AI UGC ads cost-effective create transparency obligations: viewers cannot always tell a synthetic presenter from a real one, and both Meta and TikTok have formalised policies that require some form of disclosure.

Understanding these policies is important for two reasons. First, non-compliant ads can be rejected, have reach throttled, or result in account penalties. Second, platform enforcement is increasingly automated in 2026.

Important disclaimer: This article summarises platform policies as publicly documented. It is not legal advice. Policy language changes without notice. Compliance decisions with significant legal or financial consequences should be reviewed by a qualified legal professional using the current official policy texts linked throughout.

Meta’s AI-Generated Ad Disclosure Policy

The “AI Info” Label and “About This Ad”

Meta introduced the “AI Info” labelling system for ads through a phased rollout beginning in early 2025 and expanding to third-party AI tool detection from June 2026. The label appears in the three-dot menu of an ad under “About this ad”. For ads featuring AI-generated photorealistic humans, the label may also appear next to the “Sponsored” (or “Ad”) label at the top of the ad unit itself.

Source: Meta Business Help Center — About AI info on ads created or edited with generative AI tools and Meta Newsroom, February 2025: Expanding GenAI Transparency for Meta’s Ads Products.

What Triggers the AI Info Label on Meta

Meta applies the AI Info label through two distinct mechanisms:

1. Meta’s own generative AI creative tools: When an advertiser uses Meta’s in-platform generative AI features (Background Generation, Image Generation, Add Animation), the label is applied automatically if the edit is significant. Minor changes such as image resizing or colour correction do not trigger the label. When the output includes an AI-generated photorealistic human, the label appears more prominently alongside the ad unit (not only behind the three-dot menu).
2. Third-party AI tool detection (from June 2026): Meta began using automated detection technology to identify ad creative made or significantly edited with third-party AI tools (such as Adobe Photoshop generative fill, DALL-E, Midjourney, and AI UGC platforms). When detected via industry-standard signals including C2PA Content Credentials metadata, an AI Info label is applied automatically. This requires no action from the advertiser but also means advertisers cannot prevent the label by stripping metadata.

Source: Meta Help Center — How AI-generated images in ads are identified and labeled on Meta.

Self-Certification Requirements (Political Ads)

For ads about social issues, elections, or politics, Meta requires mandatory self-certification in Ads Manager. Advertisers must disclose when the ad contains a photorealistic image or video, or realistic-sounding audio, that was created or edited using third-party generative AI tools to:

• Depict a real person as saying or doing something they did not say or do
• Depict a realistic-looking person that does not exist, or a realistic-looking event that did not happen
• Alter footage of a real event that happened
• Depict a realistic event that allegedly occurred but is not a true image, video, or audio recording

Failure to self-certify when required results in ad rejection. Repeated failures may result in account penalties.

Source: Meta Transparency Center — Ads about Social Issues, Elections or Politics.

What Does Not Trigger the AI Info Label on Meta

Not all AI involvement in ad production triggers disclosure. Meta’s policy explicitly excludes:

• Minor changes or enhancements such as image resizing or colour correction
• AI-assisted copy suggestions (text generation) where the AI suggests options but the advertiser selects and edits the final text
• Standard post-production tools that have always existed in creative workflows (cropping, brightness adjustment, noise reduction)

The distinction Meta draws is between AI-assisted (human creative process aided by AI) and AI-generated (AI systems generating primary visual or audio content).

TikTok’s AIGC Label Policy for Advertisers

The AIGC Label: Requirement and Mechanism

TikTok’s AIGC (AI-Generated Content) label requirement is set out in its Community Guidelines under “Edited Media and AI-Generated Content” (effective September 13, 2025). TikTok’s current guidelines display a further update date of September 24, 2026; because that date is still upcoming at this article’s September 17 review, verify the live policy before publishing. The requirement applies to creators and advertisers alike: content that is completely generated or significantly edited by AI and that shows realistic-appearing scenes or people must be labelled.

Source: TikTok Community Guidelines — Integrity and Authenticity and TikTok Support — AI-Generated Content.

What Triggers the AIGC Label on TikTok

TikTok requires disclosure when content contains:

• A photorealistic AI-generated or face-swapped human who looks like a real person
• AI-generated voice or audio that impersonates a real person
• AI-generated realistic scenes that could lead a viewer to believe an event took place in the real world
• Significant alterations to a real person’s appearance using AI (face replacement, body modification that changes recognition)

For AI UGC ads specifically — videos featuring synthetic presenters generated by tools like Arcads, Creatify, or MakeUGC — the AIGC label is required because these involve AI-generated photorealistic humans speaking to camera. This applies to paid ad creative submitted through TikTok Ads Manager as well as organic branded content posts.

How to Apply the AIGC Label on TikTok

TikTok offers multiple routes for disclosure:

1. Creator AIGC label (in-app): When uploading, creators can select “Your content is AI-generated” in the post settings. This adds a “creator labeled as AI-generated” indicator visible to viewers.
2. Caption, hashtag, or sticker: Creators can add a clear text caption, hashtag sticker, or on-screen watermark disclosing that the content is AI-generated.
3. Auto-labelling via C2PA Content Credentials: TikTok became the first video platform to implement C2PA Content Credentials in January 2025. When AI-generated content is uploaded with embedded C2PA metadata (as exported by tools that embed these credentials), TikTok automatically applies an “AI-generated” label without creator action. As of mid-2026, TikTok has also added invisible watermarking to content created with TikTok’s own AI tools.

Source: TikTok Newsroom — AI transparency and literacy.

What Does Not Require the AIGC Label on TikTok

TikTok’s guidelines state disclosure is not needed when:

• Making small edits such as colour correction, reframing, or cropping
• Using artistic styles such as anime, cartoon, or abstract AI art (non-photorealistic)
• Using generic text-to-speech (TTS) narration when the voice is not a recognisable imitation of a specific known individual

The defining test is whether a reasonable viewer would believe the person shown is real, or that the depicted event took place in the real world. Abstract AI art that is clearly stylised falls outside the label requirement. A photorealistic avatar presenting a product review does not.

Side-by-Side Policy Comparison

Content Prohibited Regardless of Disclosure

On both platforms, certain AI-generated content is prohibited outright — the AIGC label or AI Info disclosure does not make prohibited content compliant.

Meta

• AI-generated content in political ads that falsely depicts a real candidate as saying or doing something they did not, when self-certification is not completed
• Deepfake content that violates Community Standards on harassment, misinformation, or impersonation regardless of label
• AI-generated ads targeting people under 18 with certain content categories

Source: Meta Community Standards.

TikTok

TikTok explicitly prohibits AI-generated content that:

• Uses the likeness of any person under 18
• Uses the likeness of adult private figures without their explicit consent (a private figure is any adult who is not a public figure)
• Depicts fake authoritative sources or crisis events (fabricated news broadcasts, staged disasters or emergencies)
• Shows a public figure being bullied, making a product endorsement, or taking political stances they have not actually taken
• Is part of a coordinated influence operation or inauthentic engagement scheme

Source: TikTok Community Guidelines — Edited Media and AI-Generated Content.

For AI UGC ad producers: using a fictional AI avatar (not based on any real person’s likeness) presenting a product is generally within policy on both platforms when properly labelled. Using a real person’s face or voice without consent, or depicting a real person endorsing a product they have not endorsed, is prohibited on both platforms.

Practical Steps for AI UGC Ad Producers

Based on the current publicly documented policies on each platform, the following workflow applies to creators running AI-generated video ads on Meta and TikTok:

1. Confirm your AI UGC tool embeds C2PA metadata. If your tool (e.g., Arcads, Creatify) exports videos with C2PA Content Credentials embedded, both platforms can auto-detect and label them without additional steps. Check your tool’s export settings or support documentation to confirm whether it embeds C2PA metadata. This is increasingly common but not universal.
2. For TikTok: apply the AIGC label at upload. Whether or not C2PA metadata is present, manually applying the creator AIGC label in TikTok’s post settings ensures compliance for photorealistic AI UGC content. You can also add a visible text disclosure in the caption or as a sticker. Both TikTok Ads Manager and organic upload support this.
3. For Meta: do not strip provenance metadata. Meta’s auto-detection reads C2PA metadata. Do not run video processing steps (e.g., re-encoding through tools that strip metadata) that remove provenance signals, as this could be treated as attempting to circumvent disclosure. Let the AI Info label appear — it is not penalising to your ad performance by policy.
4. For political ads on Meta: complete the self-certification step in Ads Manager. When submitting ads that contain AI-generated or significantly AI-edited content in the social issues, elections, or politics category, use Ads Manager’s AI disclosure checkbox. Non-completion results in automatic rejection.
5. Use fictional AI avatars, not licensed likenesses of real people. Ensuring your AI UGC tool is generating synthetic presenters based on fictional actors — not the likeness of identifiable real individuals — keeps the content within policy on both platforms. Verify this with your AI tool’s own terms of service regarding avatar ownership and third-party likeness rights.
6. Keep records of which tool generated each creative. If an ad is flagged or queried, being able to demonstrate the tool used, the export date, and the C2PA or metadata status supports your position in an appeal.

Limitations & How to Verify Current Policy

• Policies change without notice: Both Meta and TikTok update their ad and content policies frequently. This article reflects policy documents publicly available as of September 2026. Verify all claims against the primary sources linked throughout before making compliance decisions.
• No hands-on compliance testing: This article does not make claims about tested enforcement outcomes. Policy documents describe intent; actual enforcement behaviour may differ and evolve.
• Not legal advice: Nothing in this article constitutes legal advice. If you are managing AI-generated advertising at scale — particularly in regulated categories, for political advertisers, or in markets subject to the EU AI Act — consult a qualified legal professional who can review the current applicable policies and laws in your jurisdiction.
• Platform-level policy vs. law: Platform AI disclosure policies exist alongside, and do not substitute for, applicable laws (e.g., FTC Endorsement Guides in the US, EU AI Act Article 50, state-level AI disclosure laws). Compliance with a platform policy does not guarantee compliance with applicable law.

To verify current policy:

• Meta AI ad labelling: Meta Business Help Center — About AI info on ads and Meta Advertising Standards
• Meta SIEP (political) self-certification: Meta Transparency Center — SIEP Ads
• TikTok AIGC label: TikTok Support — AI-Generated Content
• TikTok Community Guidelines: Integrity and Authenticity

Frequently Asked Questions