Beyond Synthetic Performers: Pennsylvania Broadens AI Ad Disclosure Requirements

Artificial intelligence (AI) is still dominating the headlines, even in a consequential election year. Anthropic’s closely watched IPO, reports of AI systems going rogue and the politics of the data centers behind a seemingly insatiable demand for AI have kept AI in front of policymakers, business leaders, lawyers and consumers alike. State legislatures have noticed, and advertising has become a particular focus.

On October 22, 2026, Governor Josh Shapiro signed , the Artificial Intelligence Deceptive Advertising Act (the “PA AI Disclosure Act”). With it, Pennsylvania joins and in requiring AI-related disclosures in advertising. The law takes effect 60 days after enactment, on December 21, 2026.

Pennsylvania’s approach differs from the New York and California synthetic performer laws in two important ways. First, those laws apply only to “synthetic performers,” meaning digitally created, human-like performers who are not recognizable as any identifiable natural person. Pennsylvania instead covers any “synthetic content,” which it defines as information, “including an image, video, audio clip or text, that has been significantly modified or generated by artificial intelligence.” Second, a disclosure is required only when synthetic content is used “in a manner that creates a false consumer perception about the nature, origin or authenticity of the goods or services being advertised.”

What the Law Requires

Under the PA AI Disclosure Act, using AI in an ad does not, on its own, require a disclosure. Its “false consumer perception” trigger echoes the deception standard under Section 5 of the FTC Act, which asks whether a representation, omission or practice is likely to mislead a consumer acting reasonably under the circumstances and whether it is material. The Pennsylvania trigger resembles the “likely to mislead” element. As a result, advertisers should expect familiar questions about the net impression an ad conveys and whether the AI-generated content misrepresents the product. Unlike the FTC standard, however, the statute does not expressly require that the false perception be material, meaning likely to affect a consumer’s conduct or decision regarding the product.

Nor does the statute have a knowledge requirement. An earlier version applied only to advertisers acting “knowingly, recklessly or willfully,” but that language was struck before passage. That sets Pennsylvania apart from New York where the disclosure obligation applies only when the advertiser has “actual knowledge” that a synthetic performer is in the ad. California likewise has no knowledge qualifier.

Is materiality relevant? Without a materiality requirement, the law could be read to require disclosure even when AI content creates a false impression that would not affect a purchasing decision. In practice, the gap may be narrow. The law reaches only false perceptions about the “nature, origin or authenticity” of the goods or services, which are the kinds of product attributes that usually matter to consumers.

California, which also has no express materiality requirement, limits its law differently: disclosure is required only when a synthetic performer is used “prominently.” The statute defines that term as being “[i]n the foreground and demonstrating or illustrating the product or service,” “[p]roviding or voicing the on- or off-camera narration or commercial message,” or “[i]llustrating or reacting to the on- or off-camera narration or commercial message.” Background or incidental figures probably will not trigger a disclosure.

In effect, both laws limit disclosure to AI content that is central to how consumers understand the product. Advertisers can, therefore, reasonably argue that an AI-generated background scene that only sets a mood falls outside the PA AI Disclosure Act, much as it would fall outside California’s prominence requirement.

New York is more categorical. Neither materiality nor consumer perception plays a role, and once the advertiser has actual knowledge, New York arguably requires disclosure whenever a synthetic performer appears in an ad.

Disclosure alone may not be enough. Pennsylvania makes this point expressly: disclosing AI use “shall not be construed to limit any recourse under law.” But the result would be the same without that language. Neither New York nor California provides a safe harbor for an AI disclosure when the ad is otherwise deceptive and violates consumer protection laws. New York’s law states that it does not “limit, reduce, or enlarge any rights any person may have” under the civil rights law “or under any other law.” California treats violations of its synthetic performer law as violations of its existing false advertising and unfair competition statutes. Under all three laws, the AI disclosure supplements general false advertising law rather than replacing it. An AI-enhanced image that overstates a product’s performance may still violate federal and state false advertising laws, even with a compliant disclosure, and substantiation obligations still apply.

Exceptions. The PA AI Disclosure Act’s disclosure requirement does not apply to content that is (1) “evidently fictional based on the content or context of the depiction,” (2) “parody, satire, commentary or criticism,” or (3) “limited to text.” Although “synthetic content” includes text, text-only content is therefore exempt, leaving AI-generated or AI-modified images, video and audio as the law’s focus.

New York’s exceptions are different and, in some respects, broader. It does not apply to audio advertisements. It also excludes ads where the use of AI “solely involves the language translation of a human performer.” Ads and promotional materials for expressive works, such as films, TV shows, streaming content and video games, are exempt as long as the synthetic performer is used in the ad the same way it is used in the work. New York also exempts the media outlets and platforms that publish or distribute a noncompliant ad, including newspapers, broadcasters, streaming services and billboards.

California, unlike New York, covers audio, so radio, podcast and streaming audio ads that use an AI-generated voice are within its scope. Its main limit is the prominence requirement rather than content-based exceptions. Like New York, California exempts platforms and publishers that merely host ads, but only until a court finds an ad unlawful and the platform is served with the order.

How to Disclose

When a disclosure is required, it must state that the ad “contains content that was modified or generated using artificial intelligence.” The disclosure must meet three requirements:

• It must be displayed in the first instance whenever the synthetic content is presented to the consumer.
• It must be presented in the same medium as the synthetic content.
• To the extent technically feasible, it must be permanent or extraordinarily difficult to remove from the synthetic content.

The statute also defines “clear and conspicuous” in detail. A disclosure must be readily noticeable, readable and understandable. It must not contradict other information in the ad, and it must appear close to the information it modifies. The statute sets specific standards for audio, visual and print ads, including point-of-sale displays and brochures. The disclosure must also be presented in the language the consumer uses to interact with or understand the synthetic content.

Like California, and unlike New York, Pennsylvania specifies what the disclosure must say. California accepts language “substantially similar” to “this performance features a synthetic performer” or “no human performer is depicted.” Pennsylvania’s required statement, by contrast, must refer to content modified or generated using AI and is not limited to performers. National campaigns using AI-generated performers may need disclosure language, or a combination of statements, that satisfies both states.

Enforcement

In Pennsylvania, enforcement rests with the government. The Attorney General or the appropriate district attorney may seek a temporary or permanent injunction against a violation of the PA AI Disclosure Act or may bring a civil action. Courts may award a civil fine of up to $3,000 plus actual attorney fees and court costs. The statute does not provide for a private right of action.

New York imposes civil penalties of $1,000 for a first violation and $5,000 for each later violation. Because the statute neither provides an express private right of action nor designates an enforcer, the New York Attorney General will likely be the primary enforcer. California’s approach, by contrast, exposes advertisers to private class action risk under its false advertising and unfair competition statutes, although consumer plaintiffs must show they suffered damages.

Conclusion

Together, the New York, California and Pennsylvania laws form an emerging state-level framework for AI disclosure in advertising. Advertisers running national campaigns will need to satisfy all three, each with its own trigger, required wording, exceptions and enforcement model.

Pennsylvania also signals a shift in the trend. The New York and California laws, both sponsored by SAG-AFTRA, ask a narrow question: does a fictional, human-like performer appear in the ad? Pennsylvania asks a broader one that false advertising lawyers will recognize: does any AI-generated or AI-modified content mislead consumers about the product? Future laws may follow either model or combine them.

More states are likely to act. The synthetic performer framework moved through two legislatures in ten months, and similar bills, probably drafted from the same template, may follow in other states. Pennsylvania now offers a second, deception-based template. Advertisers should build all three frameworks into their creative review processes before Pennsylvania’s law takes effect on December 21, 2026; watch for new state bills; and continue to treat AI disclosures as an addition to, not a substitute for, truthful and substantiated advertising.