Pennsylvania's AI Ad Disclosure Law: 2026 Guide

Pennsylvania's Artificial Intelligence Deceptive Advertising Act, signed into law by Governor Josh Shapiro, requires any ad that has been "significantly modified or generated by artificial intelligence" to carry a disclosure label in the same format as the ad itself, according to The Verge. That means an on-screen label for video, a spoken disclosure for audio, and a text label for images and written copy. If your ads reach Pennsylvania audiences in 2026, you need a plan for labeling AI-assisted creative now, not after the first complaint lands.

Key takeaways

• Pennsylvania Gov. Josh Shapiro signed the Artificial Intelligence Deceptive Advertising Act, requiring disclosure labels on AI-generated or AI-modified ads (The Verge).
• The law covers four ad formats: still images, video, audio, and text, and each requires a label matching that same format.
• The trigger is an ad being "significantly modified or generated" by AI, a threshold the law does not spell out with a specific percentage or rule.
• Businesses running AI-written copy, AI voiceovers, AI-generated images, or AI video that reach Pennsylvania audiences fall within scope, even if the business itself is based elsewhere.
• More states are expected to introduce similar AI ad labeling rules in 2026, so treating Pennsylvania's law as a preview rather than an isolated case is the smarter bet.

What Did Pennsylvania's New AI Ad Law Actually Do?

Pennsylvania became one of the first states to pass a law specifically requiring disclosure labels on AI-generated advertising. Governor Shapiro signed the Artificial Intelligence Deceptive Advertising Act, which requires any ad that has been significantly modified or generated by AI to include a label, and that label has to match the format of the ad itself, according to The Verge's reporting on the signing.

An AI-generated ad disclosure requirement is a legal rule that forces advertisers to label ads created or substantially altered by artificial intelligence, using a label format that matches the ad's own format.

That last part matters. This isn't a one-size-fits-all "AI-generated" badge stapled onto everything. A TV commercial needs an on-screen label. A radio spot needs a spoken disclosure. A text ad needs a text disclosure. The format of the label has to mirror the format of the ad.

Which Types of Ads Does the Disclosure Rule Cover?

The law covers four ad formats directly: still images, video, audio, and text. Here's how the disclosure requirement likely plays out across each one, based on the format-matching rule described in the law.

Ad Format	Disclosure Method Required	Example
Still image	Text label on the image	A social graphic with an AI-generated product photo needs a visible label on the image
Video	On-screen label in the video itself	A TV commercial with an AI avatar or AI-generated scene needs a label during that segment
Audio	Spoken or audible disclosure	A radio ad using an AI voice clone needs a spoken disclosure in the audio
Text	Text disclosure	AI-written ad copy or sponsored text content needs a written label

This is a wider net than most people assume when they hear "AI ad law." It's not just deepfake video. AI-written headlines and AI-generated product shots both fall under the same disclosure logic.

What Counts as "Significantly Modified or Generated by AI"?

The law doesn't define a hard percentage or technical threshold for "significantly modified," based on the available reporting, which leaves room for interpretation. A business using AI to crop an image or adjust color balance is almost certainly in a different category than a business using AI to generate a spokesperson who doesn't exist.

If an AI tool created the thing a viewer is responding to, the voice, the face, the product shot, the claim, that's the part most likely to need a label. Minor technical edits are a gray zone the law hasn't cleared up yet.

Until state guidance or enforcement cases clarify the line, the safer approach is to label anything where AI did meaningful creative or generative work, not just light post-processing.

Why Did Pennsylvania Pass This Law Now?

Pennsylvania passed this law in response to growing concern that AI-generated content in advertising can mislead consumers about what's real. The broader cultural shift toward labeling or flagging AI involvement is showing up outside advertising too. BBC's Blue Planet III deliberately kept imperfect, less-polished shots in its nature footage specifically to signal authenticity in an era where flawless footage reads as a red flag for AI involvement, according to The Verge's coverage of the series.

Advertising has the same credibility problem, just with higher financial stakes. A consumer who can't tell whether a product demo, a testimonial, or a spokesperson is real isn't just confused, they're a potential deceptive-advertising complaint waiting to happen.

Does This Apply to My Business If I'm Not Based in Pennsylvania?

Most likely yes, if your ads are shown to or targeted at Pennsylvania audiences, regardless of where your business is headquartered. State advertising and consumer protection laws typically apply based on where the ad is seen, not where the advertiser sits. If you run geo-targeted digital ads, radio buys, or any campaign that touches Pennsylvania zip codes, assume the disclosure requirement applies to that slice of your campaign.

This is exactly the kind of compliance layer we build into AI-managed ad campaigns for clients, matching creative review with disclosure requirements before anything goes live, get in touch if you want that audit done before your next launch.

How Should You Prepare Your Ad Campaigns for AI Disclosure Rules in 2026?

You prepare by auditing what's already running, deciding where AI involvement crosses into "significant," and building a repeatable labeling habit into your ad workflow. Here's a simple sequence that works whether you manage ads yourself or have an agency doing it:

1. Inventory your active creative. List every ad asset currently running by format: image, video, audio, text.
2. Flag AI involvement per asset. Note which assets used AI for copy, voice, imagery, or video generation versus which are fully human-made.
3. Classify minor edit vs. significant generation. A color correction is different from an AI-generated spokesperson. Be honest about where your creative actually falls.
4. Build a label template per format. Have a ready-made text label, on-screen graphic, and audio disclosure line so you're not writing one from scratch every time.
5. Document which tool touched what. Keep a simple log of which AI tool was used, on which asset, and when. This becomes your compliance record if anyone ever asks.

Most businesses skip step 5 and regret it later, because "we think it was fine" doesn't hold up the way a dated record does.

What Happens If You Don't Comply?

Enforcement details are still playing out, but violations of state advertising disclosure laws typically run through the state's existing consumer protection and deceptive trade practices framework rather than creating a brand-new penalty system from scratch. In practice, that usually means complaints can trigger investigation by the state attorney general's office, and non-compliant advertisers risk the same kind of exposure they'd face for any other deceptive advertising claim.

The bigger risk for most small and mid-size businesses isn't a fine, it's reputational. A customer who finds out your "testimonial" was AI-generated and unlabeled is a customer who stops trusting your ads altogether.

Will Other States Follow Pennsylvania's Lead?

It's reasonable to expect more states to introduce similar AI ad disclosure bills given how fast AI-generated creative has spread into mainstream advertising. Pennsylvania didn't pass this in a vacuum, it reflects a broader regulatory mood around AI transparency that's shown up in other contexts, from AI-generated political content to AI's role in search and publishing. If you run ads nationally, treat Pennsylvania's law as the first domino, not the last one.

How Does This Affect AI-Managed Advertising Specifically?

If an AI system is generating or optimizing your ad creative, that same system needs disclosure logic built into its workflow, not bolted on afterward. Businesses using AI-managed advertising should ask their provider directly: does the system flag AI-generated assets automatically, and does it apply the correct label format per channel?

This isn't a reason to avoid AI in your ad stack. AI-generated creative still tests well and often converts better when it's optimized correctly. It just means the compliance step has to be part of the campaign build, the same way ad copy review or brand approval already is.

For businesses running custom AI agents that touch marketing operations, this is also a good moment to build a standing rule: any agent that generates or edits customer-facing creative gets a compliance check before publishing, every time, no exceptions.

Worked Example: What This Looks Like for a Small Local Business

Say a home services company runs a $3,000/month ad budget across Facebook, Google, and local radio, and uses AI tools to generate some product images and write ad copy variations. Under Pennsylvania's rule, the AI-written copy running as text ads needs a text disclosure, and if any AI-generated imagery counts as "significant," it needs a visible label too. The radio spot, if it uses a fully human voice reading human-written copy, likely doesn't trigger the rule at all.

The fix here isn't expensive: a one-time audit of current creative (maybe an hour of work), a label template built once, and a 5-minute compliance check added to the campaign launch checklist going forward. That's a far cheaper fix than dealing with a consumer complaint after the fact.

FAQ Quick Reference Checklist

• Does the asset use AI-generated voice, video, image, or copy? If yes, assess significance.
• Is the AI's role significant (core content) or minor (light editing)? Label if significant.
• Does the label format match the ad format? Text for text, on-screen for video, spoken for audio.
• Is there a dated record of which AI tool touched this asset?
• Does this ad reach Pennsylvania audiences, even through geo-targeting?

If you answered yes to the first and last questions, you likely need a label, even if your business has nothing to do with Pennsylvania on paper.

Automation Atlas designs and manages AI-powered ad campaigns, and compliance checks like this are part of how we build them from day one. If you want your 2026 ad campaigns audited for AI disclosure risk before you spend another dollar, get in touch and we'll walk through your current creative with you.