A New AI Disclosure Law from Pennsylvania. But What Does It Actually Cover?

Pennsylvania Governor Josh Shapiro just signed the Artificial Intelligence Deceptive Advertising Act (SB 806). It follows similar laws in New York and California, which concern disclosing the use of “synthetic performers.” Like those laws, it seems to be aimed at transparency. However, as drafted, it leaves open a basic question: what actually triggers a disclosure?

The law states that an advertiser must disclose AI use 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.” That raises at least three issues:

• It’s unclear what “false” means. One reading is that a disclosure is required when AI is used to depict a product. Another reading is that the ad must use AI in a way that otherwise misrepresents the product, beyond the bare fact that AI was used to depict the product. But an ad that misrepresents a product is likely already unlawful under Pennsylvania’s existing false advertising law, as well as under Section 5 of the FTC Act, regardless of whether AI is used, and a label saying AI was used wouldn’t fix it.

• A disclosure is not a defense. The law expressly preserves all other legal remedies. So, an ad that creates a false perception appears to be exposed whether it carries the disclosure.

• The lead sponsor appears to read it broadly, but the text doesn’t clearly suggest such a broad reading. On the Senate floor, Sen. Nick Pisciottano said Pennsylvanians “deserve to know whether the ad they are hearing or viewing was created by a person or generated by AI,” and that the bill “will require an advertiser using synthetic content in their advertisements to clearly disclose the use of artificial intelligence.” His press release gives a similarly broad example: an AI-generated image “that could lead consumers to believe they are seeing an actual product.” The enacted text, however, requires disclosure only when the content creates a false consumer perception. A co-sponsor described the standard more narrowly, as aimed at AI content that makes a product “appear better than it is.” When another senator objected that the bill “does not tell us false as compared to what,” neither chamber amended the language. 

What is clear:

• The law covers “synthetic content,” not just “synthetic performers.” Up to now, we’ve been primarily concerned with people appearing in ads. This law extends to images, video and audio that have been “significantly modified or generated by artificial intelligence.” 

• Disclosure language and placement. The disclosure must say that the ad “contains content that was modified or generated using artificial intelligence.” It must be clear and conspicuous, appear the first time the content is presented and appear in the same medium as the content. Where technically feasible, it must also be permanent or extraordinarily difficult to remove. 

• No knowledge requirement. Language limiting the law to advertisers who act “knowingly, recklessly or willfully” was removed on the Senate floor. 

• Exemptions. The law exempts content that is evidently fictional; parody, satire, commentary or criticism; and content limited to text. 

• Enforcement. Only the attorney general and district attorneys can enforce the law. They can seek an injunction or bring a civil action, and a court may impose a fine of up to $3,000 plus attorney fees and costs. There is no private right of action.

• Effective date. The law takes effect December 21, 2026, 60 days after signing. That's a short window to implement a vague law with multiple interpretations.

How it compares to California and New York, as best we can tell:

• It isn’t limited to synthetic performers, but it may cover them in some cases. New York and California require disclosure of synthetic performers, meaning digitally created people who are not recognizable as any real person. Pennsylvania doesn’t mention performers. Instead, it applies to images, video and audio that have been significantly modified or generated by AI, but only when used in a way that creates a false perception. As a result, an AI spokesperson delivering accurate information may not require disclosure in Pennsylvania. An AI-altered real person giving a fake endorsement, which New York and California don’t reach, may.

• Knowledge. New York requires actual knowledge that an ad contains a synthetic performer. Pennsylvania has no knowledge requirement. 

• Exemptions differ. Unlike New York and California, Pennsylvania has no express exemption for ads for expressive works or for translations. 

• Enforcement. Like New York, and unlike California, Pennsylvania has no private enforcement.

Takeaways:

• Don’t expect clarity soon. Without guidance or enforcement actions, advertisers will have to decide for themselves what triggers the law. The lead sponsor’s comments suggest that enforcers may treat it as a general AI disclosure requirement, even though the text is narrower. The attorney general and 67 district attorneys may read it broadly.

• When in doubt, disclose. Because the trigger is uncertain, many advertisers will likely add disclosures to AI-generated or AI-altered images, video and audio in Pennsylvania ads rather than try to apply the test.

• Expect more states to follow. Three states have now taken three different approaches, and more are likely to come.