What happened
Pennsylvania Gov. Josh Shapiro signed the Artificial Intelligence Deceptive Advertising Act into law on Thursday, according to The Verge. The bill was introduced by State Sen. Nick Pisciottano, whose office did not immediately return a request for comment.
The law requires a label on any ad that has been significantly modified or generated by artificial intelligence. The label must appear in the same format as the ad itself. That covers still images, video, audio and text alike.
The rule applies broadly. It reaches TV commercials from large companies and small local ads alike, including a restaurant flyer for a wing special. Pennsylvania is one of the first states to require disclosures on AI content, The Verge reports.
Why it matters
AI tools can now produce a polished photo, voiceover or video in minutes, and a viewer often has no way to tell what is real. A label in the same format as the ad is meant to close that gap. A fake-looking photo gets a visible label, and an AI voiceover gets an audible one, rather than a footnote nobody sees.
The law also reaches beyond Pennsylvania's borders in practice. Labels are required regardless of where an ad was created, so a commercial made for a national run could carry the disclosure in other states too if it is also expected to air in Pennsylvania. Advertisers that run national campaigns may find it simpler to label everywhere than to cut a separate Pennsylvania version.
The move lands in a wider US fight over how AI should be disclosed and controlled. Washington is still wrestling with AI rules, and states are moving first in several areas. Readers can see that pattern in earlier RoboVez coverage of [Trump's push on AI data centres](/article/trump-ai-data-centres-dividends-midterms) and [Google's change to how Gmail data informs ads](/article/google-privacy-change-gmail-data-ads). AI-generated material is also flooding the web, as covered in [our report on AI agents and the sites fighting back](/article/ai-agents-are-flooding-the-internet-and-websites-are-fighting-back).
What it means for the US
For American consumers, the practical effect is simple: ads in Pennsylvania should start telling you when AI made or substantially changed what you are seeing. That matters most where an ad could mislead a buyer about a product, a price or a service.
For national advertisers, Pennsylvania becomes a test case. If the labelling rule works without a major legal fight, other states may copy it, and a patchwork of state rules could push big brands toward one national standard. If it stalls in court or proves hard to enforce, that will shape how other states write their own versions.
The loopholes
The Verge flags several gaps in the law as written. The phrase "significantly modified" is subjective, and the law does not draw a bright line between a light AI touch-up and a fully synthetic ad.
There are carve-outs for parody, satire, commentary or criticism, and for anything that is "evidently fictional." What counts as evidently fictional is unclear. The wording also suggests that only content that uses synthetic material in a way that creates a false consumer perception about a product or service would need a label. That leaves open questions about what counts as false, and who gets to decide.
What happens next
Watch for the details the first version of the law leaves open: how the rule will be enforced, what penalties apply, and when advertisers must comply. Those answers will decide whether the label becomes a real check on misleading AI ads or a line that companies can argue around.
Watch, too, for challenges. Advertisers may test the carve-outs, and regulators in other states will be watching Pennsylvania closely. The first enforcement cases, if they come, will set the standard for what "significantly modified" actually means in practice.
