Pennsylvania’s AI ad disclosure rule will reach radio voices on Dec. 21

Pennsylvania’s Artificial Intelligence Deceptive Advertising Act takes effect Dec. 21, requiring a clear disclosure when AI-generated or AI-modified advertising creates a false impression about the nature, origin or authenticity of what is being sold. For radio, podcast and streaming-audio spots, the notice must be audible and comprehensible to listeners, rather than buried in a rapid end-of-ad disclaimer.

The measure, Senate Bill 806, was signed by Gov. Josh Shapiro on Oct. 22 and became Act 99 of 2026. Its disclosure requirement is targeted, not a blanket label for every commercial made with AI: a synthetic spokesperson delivering accurate information may fall outside the rule if it does not mislead consumers about the product or service.

The act defines synthetic content broadly to include AI-generated or significantly modified images, video, audio clips and text. A cloned voice, a generated voice recording or a human recording heavily reworked by AI can therefore be covered. Where a disclosure is required, it must say that the material “contains content that was modified or generated using artificial intelligence,” appear when the synthetic content is first presented, and use the same medium.

That makes a text crawl inadequate for a radio advertisement. Audio disclosures must be given at a volume and pace listeners can hear and follow; where technically possible, disclosures should be permanent or difficult to remove.

The law is directed at deceptive commercial claims, such as an AI voice delivering a fabricated testimonial or a synthetic demonstration that misrepresents a product’s performance. It can also reach an AI-altered real person offering a false endorsement. It does not ban AI voices or give voice actors a right to prevent clients from using them.

Exemptions cover evidently fictional material, parody, satire, commentary, criticism and text-only content. Political advertisements are not covered, according to state Sen. Nick Pisciottano. But the statute does not precisely define “evidently fictional,” and lawyers have identified uncertainty over what constitutes a false consumer perception, how extensively content must be altered to trigger disclosure, and whether routine synthetic voices require a label.

The Pennsylvania attorney general and district attorneys can enforce the act, seeking injunctions, civil penalties of up to $3,000, attorneys’ fees and court costs. There is no private right of action for listeners. The law does not expressly exempt broadcasters distributing commercials made by outside networks, syndicators or advertisers, and it does not specify whether penalties are assessed per advertisement, per airing or by another method.

SB 806 passed the state Senate 48-2 and the House 163-40 after unanimous committee approval in both chambers. The law adds to the state-level patchwork for AI advertising, which may complicate campaigns run across jurisdictions; a disclosure also does not displace existing false-advertising rules.

California’s separate synthetic-performer advertising law takes effect Jan. 1, as we reported in September. Kelley Drye said Pennsylvania’s approach is more explicitly focused on consumer protection than the California and New York measures it compared, while Pennsylvania is described as clearly reaching radio advertisements.

The FTC has separately identified AI-enabled voice cloning as a risk for fraud, scams and biometric-data misuse, while warning that no single technical measure can prevent all harms. Pennsylvania’s new obligation begins during the holiday advertising season, leaving advertisers and their agencies and vendors until Dec. 21 to put disclosure controls in place.