Pennsylvania AI Ad Law Sets Fines Up to $3,000

Pennsylvania’s governor, Josh Shapiro, has signed a law that requires advertisers to label AI-generated content when it is used in a way that misleads consumers about what they are buying. Senate Bill 806, the Artificial Intelligence Deceptive Advertising Act, allows the state attorney general and district attorneys to seek civil fines of up to $3,000.

“I just signed into law a bipartisan bill that requires real transparency when it comes to AI use in advertising,” Shapiro wrote on social media on 22 October. “Pennsylvanians deserve to know if and when AI is being used in ads,” he added.

The General Assembly’s bill history records the governor’s approval on 22 October and lists the measure as Act 99 of 2026. It is part of a wider push on AI rules that TSN tracks on its AI regulation hub.

In brief

Pennsylvania’s SB 806 requires a clear AI label on adverts whose synthetic content creates a false impression of the product.
Text-only content, evident fiction, parody, satire, commentary and criticism are exempt.
The attorney general and district attorneys enforce it, with civil fines of up to $3,000. It takes effect 60 days after the 22 October signing.

What the law requires

The act covers “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”.

Under Section 4, an advertiser may not “knowingly, recklessly or willfully” create and disseminate an advert that uses synthetic content “in a manner that creates a false consumer perception about the nature, origin or authenticity of the goods or services being advertised” unless it carries a clear and conspicuous disclosure. That disclosure must state that “the advertisement contains content that was modified or generated using artificial intelligence”.

The label must also meet three placement tests. It has to appear “in the first instance whenever the synthetic content is presented”, in the same medium as the content, and, “to the extent technically feasible”, be “permanent or extraordinarily difficult to remove”.

The act spells out what “clear and conspicuous” means. Visual labels are judged on size, colour, contrast, location and how long they stay on screen. Audio labels are judged on volume and cadence. Labels must also be in the language the advert is aimed at.

Three kinds of content are exempt. These are content that is “evidently fictional” from its context, “parody, satire, commentary or criticism”, and content “limited to text”. A savings clause adds that a disclosure does not limit any other recourse available under law, so a labelled advert can still be challenged under existing rules.

Who enforces it and when

Enforcement sits with the attorney general or the relevant district attorney. Either can seek a temporary or permanent injunction and bring a civil action. A court may impose a civil fine of up to $3,000 and award attorney fees and court costs.

The law firm King Spry says the act creates no private right of action, although consumers keep the rights they already have under Pennsylvania’s consumer protection law.

The act takes effect 60 days after signing. AI Policy Desk and the law firm Loeb & Loeb both put that date at 21 December.

A bipartisan route through both chambers

The bill was introduced by Democratic state Senator Nick Pisciottano and drew Republican co-sponsors including Senators Tracy Pennycuick and Elder Vogel. According to the bill history, the Senate passed it 48-2 on 10 June 2026 and the House passed it 163-40 on 7 October.

Pisciottano called the law a “major step for consumer protection in Pennsylvania” in a statement reported by City & State Pennsylvania. He said it “sets a commonsense transparency standard so Pennsylvanians can make informed decisions about what they are buying, while still allowing businesses to take advantage of this technology”.

According to Loeb & Loeb, one senator objected during the Senate debate that the bill “does not tell us false as compared to what”, and neither chamber changed the wording.

The trigger is deception, not AI use alone

The duty to label does not apply every time an advertiser uses AI. It applies only when synthetic content creates a false impression about the nature, origin or authenticity of the product. An AI-generated background in a car advert may need no label. An AI-enhanced photo that makes a product look better than it is could.

That leaves the key judgement to prosecutors and courts. As the Senate exchange reported by Loeb & Loeb shows, lawmakers left open what “false” is measured against. Advertisers will have to decide for themselves which images cross the line until a case tests it.

A label does not cure a misleading advert

The savings clause matters as much as the labelling rule. A disclosure does not protect an advert that misleads in other ways, so existing false advertising law still applies. In practice, the label is a minimum requirement on top of rules that already exist.

Text is out, but images, video and audio are in

The text-only exemption narrows the law’s reach. The act still covers the formats where synthetic content is hardest to spot: images, video and audio. The 60-day window gives marketers little time to check their campaigns. The deepfake concerns that Ofcom weighed over Grok and the disclosure-based approach seen in South Korea’s AI layoff reporting bill show regulators reaching for transparency first. Anthropic’s draft federal AI safety bill takes a very different, risk-based route.