Google has now said in writing that an ad using AI to impersonate a public figure is banned, even when the ad says it was made with AI. If your ads use AI versions of real people, the label won't protect your account, and suspension comes without a warning.
What changed
On November 11, 2026, Google posted a short clarification in its Advertising Policies Help Center, titled "Updates to Unacceptable Business Practices (November 2026)". The notice says ads that impersonate public figures using AI are not allowed, even if they disclose the AI use. Google adds that the note "does not change the enforcement of the policy," and it sets no new effective date. PPC Land reported the notice in full.
The rule itself isn't new. Google has barred public figure impersonation since early 2024, and it now sits under Unacceptable Business Practices. Google classes these violations as egregious, meaning the account is suspended as soon as a violation is detected. There's no warning step and no grace period to pull a flagged ad.
What's new is the context. In July 2026 Google added AI labeling across its ad tools, so an ad with an AI label could look compliant at a glance. The November notice closes that reading.
What it means for you
If your ads contain no real person's face, voice or likeness, this doesn't affect you. If they do, and you don't have permission, it's now a clear problem. Google's examples include a singer's image in fake investment testimonials, an athlete falsely shown selling a weight-loss product, and AI videos of YouTubers pushing a get-rich-quick scheme. Those were already prohibited. The notice just takes away the idea that a label fixes them.
The notice is also quiet on several cases. Google hasn't said how it treats a brand with a contracted ambassador who uses AI to make new versions of an approved endorsement. It hasn't addressed parody or satire, and it hasn't said whether the clarification responds to a pattern it saw in ad review.
The cost of doing nothing is an account suspended immediately, and then an appeal that has to succeed before you're back. Google says it has cut incorrect suspensions by more than 80% and settles most appeals within a day, but it gives no absolute numbers, so the remaining error rate is unknown.
Show the numbers
| Measure | Change |
|---|---|
| Incorrect advertiser suspensions | −80% |
| Time to resolve appeals | −70% |
Source: Google, via PPC Land
What to do
- List every ad that uses AI-generated or AI-edited faces, voices or video of a real person.
- For each one, check whether you have the person's documented consent. Google's best-practice guidance points to documented consent before using a name, image, video or voice, so keep that record on file.
- Make sure every endorsement claim is accurate. If the person never said it, don't write it.
- Don't use the AI label as a defense. It tells viewers how an ad was made, not whether the person agreed to appear in it.
- Check that AI labeling is set correctly on ads that use AI. Google added the AI label setting in July 2026. The notice doesn't give a menu path, so look in your account settings or Google's help pages for the current location.
- If your ads contain no real person, no action is needed beyond your usual review.
If you're not sure which side of this line your ads sit on, I can run a free scan of your account and flag any AI likenesses worth a second look.
Scan my site