On November 11, 2026, Google clarified that ads using artificial intelligence to impersonate public figures are not allowed under its Unacceptable Business Practices policy, even when those ads disclose that they were generated or edited with AI. The notice, posted in the Advertising Policies Help Center under the title "Updates to Unacceptable Business Practices (November 2026)", sets no new enforcement date. According to Google, the change "is meant to help advertisers understand the policy and does not change the enforcement of the policy."

That enforcement is already among the harshest in Google Ads. Google's policy page classifies violations of the Unacceptable Business Practices policy as egregious, which means an account is suspended as soon as a violation is detected, with no prior warning.

In Short

Google has spelled out that an ad which uses AI to make it look as if a famous person is involved is still banned, even if the ad openly says it was made with AI. This matters to anyone running ads with AI-generated faces, voices or videos of real celebrities, athletes, politicians, business leaders or online creators, because Google treats impersonation of public figures as one of its most serious violations. Nothing about enforcement changes: accounts caught doing this are suspended immediately and without warning, as before, and an AI label does not work as a defense.

What the notice says

The notice is three sentences long. Its full operative text reads: "Google updated its Unacceptable Business Practices policy on Public-figure impersonation to clarify that ads that impersonate public figures using AI are not allowed, even if they disclose that they were generated or edited using AI. This clarification is meant to help advertisers understand the policy and does not change the enforcement of the policy." Beneath it sits the line "(Posted on November 11, 2026)".

There is no transition period and no effective date separate from the posting date. That is consistent with how Google frames the change: not a new rule, but a written answer to a question the existing rule left open. The help page also carries Google's standard language note, stating that "The English version is the official language used to enforce Google Ads policies," with translated versions offered for convenience only.

A page element points to the same question. Above the notice, a box headed "Need help? Ask our AI agent for support" offers three suggested prompts. The first is "How does the policy handle AI disclosures for public figure ads?" The other two ask what defines a public figure under the Google Ads policy and how advertisers can ensure their ads comply with the impersonation policy.

What the notice does not contain is just as relevant. Google did not publish the revised wording of the policy, a comparison with the earlier version, or an indication of which sentence on the policy page carries the clarification. It does not mention advertisers who hold a public figure's permission and use AI to create or alter that person's likeness, such as a brand working with a contracted ambassador. It does not address parody or satire. And it does not say whether the clarification responds to a specific pattern Google has seen in ad review.

What the policy already covered

Google's Unacceptable Business Practices page carries two named sub-sections, one on phishing and one on public figure impersonation, followed by guidance on fixing a suspended account. According to the page, public figures are people who are famous or well known, a category that explicitly includes government officials, entertainers, athletes, influencers and prominent business people. The examples Google lists are not exhaustive. They describe a singer's image used without permission in fake investment testimonials, a professional athlete falsely presented as sponsoring a weight-loss product, and AI-generated videos of popular YouTubers made to persuade people that the creators endorse a get-rich-quick scheme.

The AI example matters for reading the November 11 notice. An AI-generated video was already one of the scenarios the page describes as prohibited. The clarification therefore concerns a narrower point: whether a disclosure attached to such content changes the outcome. Google's answer is that it does not.

The page's best-practice guidance points advertisers toward documented consent before using a public figure's name, image, video or voice, and toward claims about endorsements that are accurate. That guidance is the closest Google comes, on the current page, to addressing licensed likenesses.

The prohibition itself is not new. Google tightened its Misrepresentation policy in early 2024 to target ads that impersonate or falsely imply affiliation with public figures, brands or organizations, with enforcement from March 2024 (April 2024 in France) and immediate, permanent suspension of offending accounts without prior warning. Public figure impersonation now sits on the Unacceptable Business Practices page rather than under the Misrepresentation heading used at the time.

How egregious enforcement works

Google Ads runs two tracks for violations that lead to suspension. In April 2025, Google clarified its account suspension documentation, defining egregious violations as those so serious that they are unlawful or pose significant harm to users. Those lead to suspension immediately, without warning. Other violations that lead to suspension come with a notification at least 7 days before the account is suspended.

Public figure impersonation sits on the first track. There is no grace period in which an advertiser can remove an ad after a first flag, which is why the scope of the policy matters to legitimate advertisers as much as to fraudsters.

Why disclosure was the open question

The clarification lands four months after Google built disclosure into its ad products. On July 9, 2026, Google added a "How this ad was made" panel in My Ad Center that indicates whether an ad was created or edited with AI, covering Search, YouTube and Discover. Ads made with Google's own generative AI tools are disclosed automatically, while advertisers using outside tools get a manual control. The same changelog entry, titled "Updates to AI labeling requirements (July 2026)", added an AI label setting across Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center and Ads Editor, and Google's documentation left compliance with the advertiser. Google can also apply labels automatically in certain conditions, and those labels cannot be removed.

The rollout came 24 days before Article 50 of the EU AI Act began to apply on August 2, 2026, with its transparency obligations for AI-generated content. Penalties in the EU reach up to 15 million euros or 3 percent of global annual turnover.

Disclosure and impersonation rules answer different questions. A label tells a viewer how an ad was produced. An impersonation rule asks whether the ad falsely suggests that a real person is involved, endorses a product or is speaking. A deepfake of a well-known investor recommending a trading platform deceives in the second sense whether or not a small AI label sits next to it. With labeling now a routine part of the Google Ads interface, an advertiser could plausibly have read a properly labeled AI likeness as compliant. The November notice removes that reading in writing.

There is also little evidence that disclosure is common where impersonation is the business model. An October 2026 study of Meta ads by Resemble AI found that only 2 of 3,379 AI-flagged ad files disclosed AI use. The study covered Meta's platforms rather than Google's, and the vendor used its own detection model, but it gives a sense of how rarely synthetic ad content carries any label at all.

The scale of the problem

Google's own reporting shows how large public figure scams became. The 2024 Ads Safety Report, released in April 2025, described a dedicated team of more than 100 experts assembled to deal with public figure impersonation, the permanent suspension of more than 700,000 offending advertiser accounts, and a 90 percent drop in reports of that type of scam ad. Across all policies, Google suspended more than 39.2 million advertiser accounts in 2024, and blocked or removed 415 million ads and suspended more than 5 million accounts for scam-related violations.

The 2025 report, published on April 16, 2026, gave larger headline numbers but no impersonation-specific count. Google blocked or removed more than 8.3 billion ads and suspended 24.9 million advertiser accounts in 2025, including 602 million ads and 4 million accounts for violations most closely associated with scams. Misrepresentation accounted for 421.5 million blocked or removed ads. How many of those involved AI-generated public figures, and how many carried an AI disclosure, is not broken out.

Precision matters because of the enforcement track. When suspension follows detection with no warning, a false positive costs a legitimate advertiser its account until an appeal succeeds. On November 13, 2025, Google said it had cut incorrect advertiser suspensions by more than 80 percent, resolved appeals 70 percent faster and settled 99 percent of advertiser appeals within 24 hours. Google gave no absolute numbers, so the size of the remaining error rate is unknown. For brands that legitimately license a celebrity's likeness and then use AI to produce variations of that endorsement, the risk is most relevant, and the new notice is silent on them.

The same problem on other platforms

Google is not alone in drawing the line. On February 26, 2026, Meta sued scam advertisers in Brazil, China and Vietnam, four of five cases involving celeb-bait, the use of famous faces to lure users, alongside cloaking, a staple of malvertising that shows reviewers a different page from the one users see. In April 2026, the Consumer Federation of America sued Meta in the Superior Court of the District of Columbia, alleging the company misled users about ad safety while profiting from scam ads.

Courts have moved as well. On September 16, 2026, the Frankfurt Regional Court ruled that Meta could not rely on the hosting exemption of the Digital Services Act in a case over fake ads and profiles impersonating the personal finance platform Finanzfluss and its co-founder Thomas Kehl, which included an AI-generated deepfake video. The court set fines of up to 250,000 euros per future violation. The judgment is not final.

Google's own video platform has built detection on the creator side. YouTube extended its likeness detection tool to a pilot group of government officials, journalists and political candidates on March 10, 2026, and in May opened it to all eligible creators aged 18 and over. Lawmakers, too, are focusing on consent rather than labels. Denmark's government on October 8, 2026 tabled a bill that would make publishers prove consent before making realistic digital imitations of a person public, from January 1, 2027 if passed. In the United States, the right of publicity already gives individuals legal control over commercial use of their name, likeness and voice.

Seen together, these developments point in one direction. Disclosure rules govern how synthetic content is presented to audiences. Consent and impersonation rules govern whether the content may be used at all. Google's notice places its ad policy firmly in the second category.

A run of clarifications

The notice is the latest in a series of Google Ads policy edits presented as clarifications rather than changes. On October 14, 2026, Google added dozens of examples to its Sexual content policy and stated that enforcement was unchanged. In November 2025, it added an example to its Circumventing Systems policy stating that false information during advertiser verification leads to suspension, again without an enforcement change. In August 2025, it reorganized the Enabling dishonest behavior policy on the same terms.

Not every recent notice has been a clarification. On October 22, 2026, Google split its child safety rules into two penalty tiers, with a new enforcement start of January 11, 2027 and different consequences for each tier. The November 11 notice belongs to the first group: it changes no deadline and no penalty.

For marketers, the practical significance lies in the interaction between two systems Google now runs in parallel. AI labeling is a compliance setting advertisers control or that Google applies. Impersonation is an egregious policy that ends accounts on detection. The clarification confirms that the first cannot be used to satisfy the second. Will reviewers treat AI-made content featuring a public figure who has agreed to it any differently? The published text does not say.

Timeline

Summary

Who: Google, through its Advertising Policies Help Center. The clarification affects any advertiser running ads that feature AI-generated or AI-edited likenesses of public figures, a category Google defines as famous or well-known people including government officials, entertainers, athletes, influencers and prominent business people.

What: Google clarified that ads using AI to impersonate public figures are not allowed under its Unacceptable Business Practices policy, even if they disclose that they were generated or edited using AI. Google says the clarification does not change enforcement. Violations remain egregious, which means account suspension upon detection and without prior warning. Google did not publish revised policy wording, and the notice does not address AI likenesses used with a public figure's consent.

When: The notice was posted on November 11, 2026. There is no separate effective date, since enforcement is unchanged; the underlying prohibition on public figure impersonation has been enforced since March 2024 (April 2024 in France).

Where: The Google Ads Advertising Policies Help Center, with the policy applying to Google Ads globally. Google states that the English version of its policies is the official language used for enforcement.

Why: Since July 2026, Google has offered AI labels and a "How this ad was made" panel to disclose AI-created ads, partly to meet the EU AI Act's transparency obligations from August 2, 2026. That raised the question of whether a disclosed AI likeness of a public figure would be treated differently. Google's answer is that disclosure describes how an ad was made but does not make impersonation acceptable, in line with the broader industry and legal focus on consent for the use of real people's likenesses.