AND AWAY WE GO: Pennsylvania Signs AI Deceptive Advertising Act. What it Means for Brands

Pennsylvania has just signed their AI Deceptive Advertising Act into law. The law so let’s cut into what it is, what it isn’t and what it means for brands placing advertisements.

What is the Pennsylvania AI Deceptive Advertising Act?

• Consumer protection regulations in the State of Pennsylvania intended to establish transparency standards for using AI in commercial advertising content
• Specifically, the Act aims to protect consumers from deceptive advertising that intentionally uses “synthetic content in a manner that creates a false consumer perception regarding the nature, origin or authenticity of the advertised goods or services.”
• Requires “clear and conspicuous” disclosure that states the advertisements contains content that was modified or generated using AI.
• Also mandates that an advertiser may not create and disseminate ads that uses synthetic content in a manner that creates false consumer perception.

How does the AI Deceptive Advertising Act Set Definitions?

• AI (including Generative AI) is a machine-based system that can (for human-defined objectives) make predictions, recommendations or decisions including real or virtual environments.
• Clear and conspicuous disclosures much be in a size, color, contrast and location with duration and audibility that is readily noticeable and understandable. Must be presented in the language of the interaction.
• Synthetic content is information, including image, video, audio or text that has been significantly modified or generated by AI.

What’s the actual requirement?

• Clear and conspicuous disclosure in the language and medium of the advertising interaction
• Displayed at first instance of presentation to the consumer
• The disclosure must be difficult to remove and cannot be obscured.

Are there exceptions to the Pennsylvania AI Deceptive Advertising Act?

• Clearly fictional content, including context that is clearly fictional (unicorn advertisers…breathe easy)
• Parody, satire, commentary or criticism
• Content limited to text

What is the consequence and who will enforce the law?

• The law will be enforced by the Pennsylvania Attorney General. There is no noted independent office or investigators noted in the bill.
• The Attorney General or a district attorney may bring action on behalf of the Commonwealth of Pennsylvania. Primary initial action will be to declare a practice unlawful and seek temporary or permanent injuction to address the method act or practice.
• The court may award damages which may include a civil fine that “does not exceed” $3,000
• This will be considered a civil offense.

So…what does this all mean?

For most brands leveraging AI to generate localized scenes or personalize creative iterations for their advertising, this is a case of adding a clearly legible notice that the images, video or asset included AI generated content. With the clarifications around assets that are clearly fictional or intended as satire, there is enough wiggle room that creatives can rest easy that they can still create to the wild brief of a unicorn driving down the streets of Scranton and getting hit by another car driven by a llama…as an example.

But, all advertisers, especially small businesses, should take note that there are NO exceptions for company size or ad buy limitations…so if you created an image for your pizza shop that shows you making pizza for the Eagles and that is NOT true…better get that disclaimer ready (among other legal defenses on why you inappropriately used the Eagles logo and the such.)

While intended to be a public safety measure to protect against flagrantly false advertising generated with AI, there are some fairly big holes in the Bill’s language that those truly intending to deceive can roll a truck right through. Take for example language in Section 4.a in which language around advertiser intent was struck out…it doesn’t matter if the creation of false consumer perception was done knowingly or by accident, the advertiser must disclose or face the consequences.

The law also don’t clearly define what a false perception is or could be, which could also lead to grey area for truly deceptive advertisers to hide. There are already measures in Pennsylvania law protecting against false advertising, regardless of method of creation…so many lawyers piping into the conversation are thinking that those FTC laws may already make the truly pernicious unlawful, so it didn’t need to be outlined here.

The questions will arise when trying to determine if the ad itself has crossed that threshold of perception, or if it specific to the product or service being advertised. Because this is a fairly broad definition of content and does not specifically all out use of synthetic actors as opposed to all synthetic content, if a synthetic actor is used that does nothing to impact or influence consumer perception of the product that synthetic actor is holding, it is unclear if that scenario requires a disclosure. In open hearings on the bill, a co-sponsor even noted that it was not as focused on what happens when AI makes the actor holding the product look better than it was on when AI actually makes the product look better than it actually is.

So…if a video ad incudes a synthetic actor talking about a new cleaning product, if all representations of the product and the results of using that product are accurate and AI is not used to manipulate the product or the results from use of that product, that ad may not require a disclosure in Pennsylvania.

For brands looking to over-index to be safe, reading this disclosure requirement for anything that has been created with AI might be safest.

How does this stack up to other legislation?

This is just the start and many other bills and acts are passing or are in committee as I write this. Laws in California and New York already require disclosure when synthetic performers are used. This Pennsylvania law is not that specific and broadens the scope to synthetic content…but only when that content creates false perception.

Using the example I used above, while in Pennsylvania the ad with the synthetic actor might not require disclosure, IT WOULD require disclosure in California and New York as it uses a synthetic actor and all synthetic performers not recognizable as any real person must be disclosed.

Illinois and Minnesota have passed laws around deepfakes, targeting the unauthorized use of AI to create synthetic replicas of real humans either living or deceased in commercial ads. Both largely fall under general protections of individual rights from publicity.

At the Federal level, there are multiple proposals in front of multiple committees around AI regulation and disclosure requirements around generative content and assets. For example, the AI Labeling Act of 2026 (*NOT to be confused with the AI Labeling Act of 2024 or 2025…not kidding…this is our reality*) proposed disclosures for AI generated content that must be both clear and conspicuous to human audiences but also be machine-readable. The bill also attempts to hold online platforms to account, noting each online platform shall ensure content that requires discloser can be clearly identified and accessible. Enforceable by the US Attorney General, this proposes a statutory damage fine of not more than $2,500 per act of circumvention.

So…

Get those disclosure notes and audio bytes ready. Have the early conversations around deceptive actions and consumer perception…and remember, perception is in the eye of the beholder. If you think it goes too far…it probably does. Make sure you bring a diverse set of eyes to the party…your individual design perception may not be the best arbiter of what is and is not deceptive manipulation.

Ask, document, declare.