Pennsylvania Sets First US State Precedent With Mandatory AI Ad Labeling Law

Pennsylvania has become the first U.S. state to pass legislation specifically requiring advertisers to disclose when artificial intelligence is used to create or significantly modify advertising materials. Governor Josh Shapiro signed the Artificial Intelligence Deceptive Advertising Act (SB 806) this week, setting a precedent that other states are likely to replicate in the coming months.

The law, which takes effect on December 21, 2026 — 60 days after signing — applies to any advertisement, whether a still image, video, audio clip, or text, that has been "significantly modified or generated by artificial intelligence" and that could "create a false consumer perception" about the nature, origin, or authenticity of the advertised goods or services. It requires a clear statement that the ad "contains content that was modified or generated using artificial intelligence," visible upon first display and in the same medium as the synthetic content itself.

What makes this law particularly notable is the gap between the strict statutory text and the broad intent expressed by its primary sponsor. Senator Nick Pisciottano (D-Allegheny) stated unequivocally on the Senate floor that Pennsylvanians "deserve to know whether the ad they are hearing or viewing was created by a person or generated by AI." The enacted text, however, triggers disclosure only when synthetic content is used "in a manner that creates a false consumer perception" — a much narrower standard than the sponsor's public framing suggests.

The central ambiguity: when must you disclose?

The most debated aspect of the legislation is the word "false." A partner at the Los Angeles law firm Loeb & Loeb LLP, who analyzed the final statutory language, identified at least three possible interpretations:

A broad reading: whenever an AI-generated image of a product is used, even if it accurately represents the product, this creates a false perception because consumers believe they are seeing the actual product when they are seeing a digital reconstruction. Under this interpretation, virtually every advertisement using AI rendering would require disclosure.

A narrow reading: disclosure is required only if the AI-generated image actually misleads consumers about something material — if the product appears better, different, or more advanced than it truly is. An AI-generated image that faithfully depicts the product, in this view, would not need a label.

A redundancy reading: if an ad uses AI to create a false representation of a product, it is already illegal under Pennsylvania's existing deceptive advertising law and Section 5 of the Federal Trade Commission Act. In this case, the new AI law would be a complement with no additional practical effect.

The statute does not resolve this ambiguity. When one senator on the Senate floor explicitly asked what "false" means "as compared to what," no amendment was offered to clarify the language.

What's excluded

The law includes several explicit exemptions: content that is "evidently fictional" (such as a science fiction movie trailer using AI-generated characters), parody, satire, commentary or criticism, and content limited to text. The text-only exemption is notable: an advertisement in text claiming "this product was made with AI" would not be covered, while an AI-generated image of the product, yes.

Enforcement authority rests exclusively with the state Attorney General and local district attorneys — there is no private right of action. Penalties are limited to a fine of up to US$ 3,000 per violation, plus attorney's fees and costs. For advertisers spending millions per campaign, the fine represents a relatively modest risk.

The national context

Pennsylvania is not the first state to address AI regulation in advertising — New York and California have already passed similar laws concerning "synthetic performers." What distinguishes Pennsylvania's law is that it covers not just people, but any "synthetic content": product images, explanatory videos, audio clips, and text generated or significantly modified by AI.

The American Advertising Association (AAPA) and the Interactive Advertising Bureau (IAB) already developed voluntary AI content labeling standards in 2025. As The Verge observed in its reporting, those standards are widely ignored. Pennsylvania's law attempts to solve this enforcement problem with legal force — but the ambiguity of the text may render enforcement so uncertain that many advertisers simply don't know whether they need to comply.

Without guidance from the Attorney General or enforcement actions in the first months, industry practice will depend on individual interpretation: some district attorneys (one for each of Pennsylvania's 67 counties) may adopt a broad reading, while others may deem the law irrelevant to ads that depict products realistically even when using AI. In an election year, political pressure for rigorous versus lenient enforcement could vary enormously from county to county.

Sources: The Verge, Loeb Law, Ginsburg Law Group

✓ Independent sources cross-checked and verified before publishing