Pennsylvania Governor Josh Shapiro signed the Artificial Intelligence Deceptive Advertising Act into law on Thursday, establishing disclosure requirements for advertisements generated or substantially altered using artificial intelligence. The measure covers advertising across still images, video, audio and text, with disclosures required in the same format as the advertisement.
The law reaches beyond advertisements produced within Pennsylvania. Its requirements apply regardless of where an advertisement was created, making the state’s rules relevant to businesses and advertising producers operating elsewhere. A campaign intended to run in Pennsylvania alongside other markets could therefore carry an AI disclosure beyond the state’s borders.
The measure’s reach spans businesses of different sizes and advertising budgets. Television campaigns from large companies and promotional materials for local businesses can both fall within its scope. The central considerations are the use of AI, the extent of its role in creating or changing the material, and the law’s provisions concerning misleading consumer impressions—not simply the size of the advertiser or the medium used.
However, the legislation leaves important questions about which advertisements must carry labels. It uses a threshold of significant AI generation or modification, a distinction that will require interpretation. Using an AI tool somewhere in the production process is not necessarily the same as producing an advertisement that meets that threshold, and the available account does not establish a precise dividing line.
A further provision concerns synthetic material used in a way that creates a false consumer perception about a product or service. That language complicates any reading of the measure as a blanket labeling requirement for every advertisement involving AI. Determining whether a disclosure is required may depend not only on how an advertisement was produced, but also on what impression its synthetic elements give consumers.
The act also includes exceptions for parody, satire, commentary and criticism, as well as material that is plainly fictional. Those categories introduce another interpretive question: when an audience should be expected to recognize that an advertisement’s content is invented. The distinction matters because advertising routinely uses imaginary settings, dramatizations and visual effects without presenting them as literal depictions of a product’s performance.
Generative AI adds to those longstanding advertising issues by enabling the creation of realistic-looking images, synthetic voices, video and promotional copy. A disclosure identifies AI’s involvement in the material; it does not, by itself, resolve whether the underlying product claims are accurate. Conversely, the use of synthetic content does not automatically establish that an advertisement makes a false claim.
State Senator Nick Pisciottano introduced the bill. With the legislation signed, its practical scope will turn on how its significant-modification threshold, consumer-perception language and fictional-content exceptions are interpreted. Those distinctions will be central to determining which advertisements need disclosures and which remain outside the requirement.
Sources: The Verge