Guide
AI Ad Disclosure Laws: What Advertisers Need to Know in 2026
Several AI disclosure rules took effect in the space of two months this summer, and advertisers are asking which of them apply to their campaigns. The short answer is that it depends on what the ad contains, who built the AI system behind it, and where the audience is. This guide summarizes three of the main measures, explains whom each one targets, and describes what they mean for conversational advertising. It is a summary for marketers, not legal advice. Check specifics with counsel.
- Published
- Oct 2026
- Reading time
- 5 min
Why AI Ad Disclosure Laws Are Arriving Now
Generative tools can now produce lifelike people, voices, and images at low cost. Lawmakers have responded by requiring that audiences be told when they are looking at synthetic content or talking to a machine. The rules differ in scope, and none of them was written specifically for an ad appearing inside a live AI conversation. That gap is covered later in this guide.
New York: The Synthetic Performer Disclosure Law
New York amended General Business Law § 396-b so that advertisements featuring a synthetic performer must carry a clear disclosure. The law took effect on June 9, 2026. A few points matter for advertisers:
- Who is covered. A business that, for a commercial purpose, produces or creates an advertisement in any medium.
- Knowledge standard. The duty applies where the advertiser has actual knowledge that a synthetic performer appears in the ad.
- Scope. It is narrow: computer-created or modified assets intended to appear as non-identifiable human performers. Ads for expressive works such as films and video games are excluded if the use is consistent with the work.
- Penalties. Reported civil penalties are $1,000 for a first violation and $5,000 for later ones. The law does not expressly create a private right of action.
California SB 942: The AI Transparency Act
California's AI Transparency Act (SB 942), as amended by AB 853, became operative on August 2, 2026. Its main duties fall on providers of generative AI systems with more than one million monthly users that are accessible in California, not on every advertiser using their output. Covered providers must offer a free public tool to detect content their system produced, give users an option to add a visible disclosure, and embed hidden provenance data in AI-generated image, video, and audio. The law applies to those media types and not to text. Providers must also pass these duties to licensees. The reported penalty is $5,000 per violation, with each day counted separately, enforced by the state Attorney General. Some further duties for large platforms start in 2027.
For advertisers, the practical effect is indirect. Expect AI image and video tools to add provenance marks, and expect platforms to ask whether your creative carries them.
EU AI Act Article 50: Transparency Obligations
Article 50 of the EU AI Act began applying on August 2, 2026. Its main requirements are:
- Chatbots. Systems that interact with people must be designed so users are told they are dealing with an AI, unless that is obvious.
- Machine-readable marking. Providers must mark AI-generated output in a detectable format. Systems already on the market before August 2 have until December 2, 2026 to comply with the marking duty, while new systems get no grace period.
- Deepfakes. Deployers must label deepfakes, and AI-generated text published to inform the public on matters of public interest in the absence of human editorial control.
The European Commission issued non-binding guidelines on Article 50 on July 20, 2026. Fines can reach €15 million or 3% of worldwide annual turnover, whichever is higher. Sources differ on some grace-period details, so verify dates against the Official Journal and Commission guidance.
Launch your first
campaign in seconds
Set up in the platform, pick your markets and a budget, and launch when you're ready. No sales call needed.
Start nowOr we set up your
campaign with you
Not sure where to start? Tell us your goals and markets, and our team plans, launches and manages it for you.
Book a callWhat AI Ad Disclosure Laws Mean for Conversational Advertising
| Rule | Mainly targets | Relevance to an ad in an AI chat |
|---|---|---|
| New York § 396-b | Advertisers whose ads include synthetic performers | Applies if your creative includes a synthetic performer, wherever it runs |
| California SB 942 | Large generative AI providers | Indirect: affects AI-made image, video, and audio creative you use |
| EU AI Act Art. 50 | Providers and deployers of AI systems | The AI assistant itself must disclose it is AI; labeling duties also touch deepfakes |
None of the three prescribes how to label a sponsored message sitting next to an AI's answer. Industry guidance fills part of the space: IAB's framework is covered in our explainer on the AI Transparency & Disclosure Framework, which notes that it does not yet offer a dedicated spec for ads inside a live conversation.
What AI Ad Disclosure Laws Do Not Settle
Three open questions are worth watching. First, who is responsible when an ad is built by an agency, produced with a third-party AI tool, and served through a platform: the laws assign duties to different parties, and contracts increasingly specify who carries what. Second, how far the rules extend to ads that are not made with AI but are shown inside an AI product. Third, how regulators will read "clear and conspicuous" for a format as new as an in-chat placement. Expect guidance, and probably enforcement examples, to answer these over time.
Other jurisdictions are also moving, and further state-level measures may already exist or follow. This overview covers only the rules above, which we verified; it should not be read as a complete list for any other state or country.
In short, the laws summarized here mostly regulate synthetic content and AI providers, and they leave the placement of ordinary sponsored messages to existing advertising rules and platform policy. That is why clear labeling of paid placements remains a matter of good practice for now, and why advertisers should watch for guidance rather than assume the question is closed.
A Practical Checklist
- Audit creative for synthetic people, voices, or digitally altered likenesses, and record what you find.
- Ask each partner how sponsored content is labeled and how sensitive conversations are excluded. Thrad's answers are on its safety page.
- Keep records of what was disclosed, where, and when.
- Disclose conservatively. Labeling when it may not be strictly required costs little; missing a requirement costs more.
- Review again every quarter. These rules are new and guidance is still being published.
For the wider context on how brands and publishers operate in this category, see our pillar guide to conversational advertising.
