Pennsylvania Will Require "AI Used" Labels on Deceptive AI Ads

Pennsylvania is making disclosure mandatory for AI advertising that misleads consumers. Governor Josh Shapiro signed the Artificial Intelligence Deceptive Advertising Act on October 22 local time. The law takes effect 60 days after approval. The core idea is simple: consumers who see an ad should be able to tell that AI was involved.

Which AI ads need a label

Whether the law applies depends on two things considered together: whether AI was used, and whether the resulting content creates a false impression for consumers. It covers ads that use AI-generated or substantially altered images, video, or audio in ways that mislead people about the nature, source, or authenticity of a product or service. The law regulates knowingly, recklessly, or intentionally creating and distributing such an ad, or causing it to be created and distributed.

There are exceptions. Obvious fiction, parody, satire, commentary, and criticism do not require a label. Text-only content is also exempt. The law’s definition of “synthetic content” (information generated or substantially altered by AI) does include text, but content consisting solely of text was carved out separately.

How the disclosure must appear

The disclosure must appear in the same medium when the AI-generated or substantially altered content is first presented to the consumer. Font size, contrast against the background, and how long the notice stays on screen all have to be sufficient for a consumer to read and understand it. For audio disclosures, volume and speaking pace must likewise be intelligible. Where technically feasible, the disclosure must be made impossible, or extremely difficult, to remove.

For violations, the state attorney general or a district attorney with jurisdiction can ask a court to halt the advertising or file a civil suit. Courts can impose civil penalties of up to $3,000 per violation. Actual attorney’s fees and court costs can also be assessed. Advertisers need to check both whether a given ad falls within scope and how the disclosure is presented.

How it differs from New York

New York’s 2026 advertising disclosure law uses a different trigger. New York addresses the use of AI “synthetic performers” that appear to be real people in ads. Pennsylvania focuses less on the type of performer and more on how the AI content in an ad misleads consumers about products and services.

Pennsylvania also saw a separate bill pass the state House in 2024 that would have required labeling of AI-generated content more broadly. This new law was advanced in connection with the state’s existing consumer protection statute, and it targets advertising that creates false impressions about products and services.

For Korean companies, the detail worth watching is the method of disclosure. The law specifies that the label must appear in the language, or by another means, that the consumer uses to understand or access the content. For consumers, this means more information with which to judge whether AI was used in an ad. The open questions in enforcement are how regulators will determine that a false impression was created, and how high the bar will be set for a disclosure that consumers can actually notice.