Pennsylvania enacts law requiring disclosure labels on AI-generated ads

What the law requires

Pennsylvania Governor Josh Shapiro has signed the Artificial Intelligence Deceptive Advertising Act, a new state law requiring disclosure labels on advertising that has been generated or substantially altered with artificial intelligence. According to The Verge, the signing took place on Thursday.

The obligation spans every major ad format. Whether an advertisement is a still image, a video, an audio spot or plain text, The Verge reports, it must include a label delivered in the same format as the ad itself. Nor does the law distinguish between large and small advertisers: it covers everything from television commercials backed by multi-billion-dollar companies down to a local restaurant's printed flyers promoting a wing special.

One of the first state mandates

Pennsylvania is among the first states in the country to require disclosures on AI content, according to The Verge. The law also attaches to where an ad runs rather than where it is made: labels are required for any advertisement shown in Pennsylvania regardless of where it was produced.

That geography-blind design could have spillover effects beyond the state's borders. A national campaign that airs in Pennsylvania may find it simpler to add disclosures everywhere than to maintain separate versions of the same commercial, meaning viewers in other states could start seeing the labels even though their own legislatures have not acted.

Where the law gets fuzzy

As The Verge observes, the statute leaves considerable room for interpretation. There is no precise definition of what makes an ad "significantly modified" by AI, leaving that judgment to vary from case to case. The law also carves out exceptions for "parody, satire, commentary or criticism" and for anything "evidently fictional" — another phrase whose boundaries are left undefined.

Perhaps most notably, The Verge reports that the wording appears to reach only synthetic content used "in a manner that creates a false consumer perception" about a product or service. That raises further unresolved questions about what qualifies as false under the law and who is empowered to make that determination.

The bill was introduced by State Senator Nick Pisciottano, whose office did not immediately respond to a request for comment from The Verge.

Why it matters

Synthetic media is now cheap and accessible enough to show up in everyday marketing, yet most jurisdictions have no rules requiring advertisers to disclose when an ad's voice, imagery or copy came from a model. Pennsylvania's move is an early attempt to close that gap at the state level, and because the requirement applies to any ad running in the state regardless of origin, a single legislature's decision could effectively set labeling practice for national campaigns.

The open questions are just as consequential. Without firm definitions of "significantly modified," "evidently fictional" or "false consumer perception," enforcement will hinge on interpretation, and advertisers, platforms and regulators may read the same advertisement very differently. How Pennsylvania applies the law in practice, and whether other states copy its approach or tighten its language, will help define the emerging baseline for AI transparency in advertising.