How to disclose AI-generated video ads
This is general information, not legal advice. Platform rules change, and the dates below are the ones shown on each page when we read it.
Meta: the AI info label
Meta's post on advertising and generative AI (first published 3 February 2025, updated 1 June 2026) says the "AI info" label applies to ads that are "created or significantly edited with our generative AI creative tools". The label sits in the "About this ad" panel. Meta also says it will begin "automatically detecting ads created or edited using third-party AI tools through industry-standard signals" and apply the same label when it does.
The post was first published on 3 February 2025 and updated on 1 June 2026. Its wording on where the label appears has changed over time, so check the current version. Source: Meta, "Expanding GenAI Transparency for Meta's Ads Products", updated 1 June 2026.
The practical point: you do not wait for Meta to spot your video. If your ad was made or edited with an AI tool, assume the label may apply, and make sure the ad does not contradict it (this is our reading, not a Meta instruction).
TikTok: an AIGC label, or your own disclosure
TikTok's misleading and false content policy for ads says that "significantly edited media and AIGC content is allowed if the following requirements are met". The first requirement is to "apply the AIGC label, or by adding a clear disclaimer, caption, watermark, or sticker of your own." The policy also says that "if we identify AI-generated content that has not been disclosed, your ad will be rejected or restricted."
The policy gives examples on both sides of the line. Counted as significant edits are content that is "completely AI-generated", a primary subject "doing something they didn't actually do", and a primary subject saying something they did not say "using AI voice-cloning". Counted as insignificant are "adjustments to lighting, brightness, or color saturation", "removing or modifying backgrounds" and denoising. The page says the policy was last updated in April 2026. Source: TikTok for Business, "Misleading and false content", last updated April 2026.
If you are unsure which side your edit falls on, the policy itself suggests using caution and labelling it.
YouTube: the altered or synthetic content question
YouTube's help page on altered or synthetic content covers creators uploading videos. It asks you to disclose realistic content that "makes a real person appear to say or do something they didn't do", "alters footage of a real event or place", or "generates a realistic scene that didn't actually occur". It also lists "creates music that's the main focus of the video".
Exempt are non-realistic content and minor edits, which the page describes as "primarily aesthetic". Its examples include "using generative AI tools to create or improve a video outline". When you answer "Yes" to AI use, YouTube adds a label, which may appear in the player or in the expanded description depending on the content. The page says "Disclosing AI content alone won't limit a video's audience or impact its eligibility to earn money." The page shows no publication or update date. Source: YouTube Help, "Altered or synthetic content", no date shown.
This page is about YouTube uploads. I have not checked Google's advertising policies for paid placements, so do that separately before running a YouTube ad.
UK: the CAP Code applies however the ad was made
The Advertising Standards Authority (ASA) and the Committee of Advertising Practice (CAP) do not have a separate AI rulebook for ads. The ASA's four-point guidance on AI and deepfakes, dated 11 June 2026, begins with media neutrality: "The Code is media-neutral, meaning the rules apply regardless of how the content was created." It adds that "the advertiser remains responsible for any output used in their ads." Source: ASA, "AI and Deepfakes: Four Things Advertisers Need to Know Before They Hit Run", 11 June 2026.
On endorsements, the same guidance says that if an AI-generated celebrity "endorsement" "looks real enough to mislead consumers, it's likely breaching the Code". An earlier ASA page dated 7 February 2025 says that "AI-generated images will breach the Code, just as any other type of image would." Source: ASA, "AI, advertising, and the policy landscape", 7 February 2025.
The ASA's pages I read did not set out a specific UK rule on disclosing AI use. A sensible question to ask is whether your audience could be misled if the AI use were not disclosed. The ASA pages above do not answer that for you. Check the current CAP guidance on that point before you rely on it. The same ASA guidance says that personality rights and intellectual property sit outside its remit, so if you are using a real person's likeness, take legal advice.
A checklist before you run an AI video ad
- Describe what the AI did. Was the whole clip generated, a face or voice swapped in, a real event altered, or only the colour and background changed? Write it down in one sentence.
- Match it to each platform's rule. The Meta, TikTok and YouTube triggers above are different. A clip can need a label on one platform and not another.
- Use the label the platform gives you. If there is a toggle for AI-generated content, switch it on. If there is no toggle and the platform allows a disclaimer, write one that a viewer can see in the ad itself, not only in the small print.
- Never fake a person. A made-up testimonial, or an AI "celebrity" who looks like they endorse the product, risks misleading viewers. The ASA says an AI-generated celebrity endorsement that looks real enough to mislead is likely to breach the CAP Code.
- Keep the records. Keep the script, the tool used and the original files. If a platform or a regulator asks how the ad was made, you can answer.
- Recheck before each campaign. Meta's post was updated in June 2026, and the TikTok policy in April 2026. Read the current version each time.
Where an AI UGC reel fits
OneTake turns a script into a captioned AI UGC reel with an AI actor. If you run one of those reels as a paid ad, the rules above apply to it in the same way they apply to any other AI-made video. The fair answer for any AI video is the same: work out what the tool changed, check the platform rule, and disclose where the rule asks you to.
Hiring a creator who is filmed on camera is a different route with its own disclosure rules, and for some brands it is the right choice. The reel format works for some businesses and not others, and the rules above are a reason to be clear with your audience, not a reason to avoid AI.