AI Disclosure and Labeling Laws by State: SB 942 (2027)

As of November 4, 2026, we found no federal statute in the public laws of the 119th Congress that requires labels on AI content, and, by our count of enacted laws and subject to the search limits shown on our AI laws tracker, at least seven states have general AI disclosure or labeling laws, meaning laws not limited to elections, to one sector or to chatbot safety: California, Connecticut, Maine, New York, Pennsylvania, Utah and Washington.

Status as of November 4, 2026: The laws of California, Connecticut, Maine, New York and Utah are in force. Pennsylvania's Act 99 takes effect about December 21, 2026, and Washington's E2SHB 1170 takes effect February 1, 2027. California's platform duties and SB 1050 and Utah's HB 276 are scheduled for January 1, 2027. In the European Union, AI Act Article 50 has applied since August 2, 2026.

This page sorts those laws by what they actually require, because "AI disclosure law" covers very different duties. Laws limited to elections, health care, companion or conversational AI chatbot safety, government agencies, employers, real estate licensees, data brokers or frontier model developers are outside the count, as are notices about automated decisions. It is general legal information, not legal advice, and it reflects what we read through November 4, 2026.

AI Disclosure and Labeling Laws by State

The table lists the seven counted states. In the public laws of the 119th Congress, we found no federal statute that requires labels on AI content, and the FCC's February 2024 declaratory ruling that AI-generated voices in calls are "artificial" under the Telephone Consumer Protection Act is a ruling on an existing statute, not an AI label. We ran AI and content-provenance terms through the full text of all 136 public laws Congress.gov listed on November 4, 2026, and checked the hits for labeling duties.

How to read this count. It rests on our state-by-state search and the limits shown on our AI laws tracker, including states where we ran only a title scan. In the disclosure column, our research did not confirm this category either way for the District of Columbia, Iowa, New Jersey, North Carolina, Ohio or Virginia. A count of "at least seven" is a floor, not a finding that the other states have nothing. Utah's SB 226 is a judgment call: it reaches only systems designed to simulate human conversation, but it is a disclosure rule for suppliers in consumer transactions, not a chatbot safety law, so we count it. Without it, Utah would count only on HB 276, and four of the seven would have a law in force.

Who the duty falls on. The same phrase, "AI disclosure," puts the duty on different people:

AI Video Rules Since August 2026: EU Article 50 and California SB 942

Two sets of rules on AI video and other synthetic content started on the same day, August 2, 2026, and they work differently. Neither is a US federal rule.

California. The AI Transparency Act became operative August 2, 2026. Its latent disclosure duty covers image, video and audio content a covered provider's system creates or alters, other than by minor modification. SB 1000 rewrote the act on September 30, 2026, as an urgency statute with immediate effect.

European Union. Under Regulation (EU) 2024/1689, the AI Act, the Article 50 transparency duties have applied since August 2, 2026. Providers of systems generating synthetic audio, image, video or text must mark outputs in a machine-readable format, with exceptions for assistive standard editing and for systems that do not substantially alter the input. Deployers must disclose deepfakes, with a lighter rule for evidently artistic or satirical work.

The Omnibus, Regulation (EU) 2026/1744, added a transition. Under Article 111(4), providers of systems generating synthetic audio, image, video or text content placed on the market before August 2, 2026 "shall take the necessary steps in order to comply with Article 50(2) by 2 December 2026." That transition covers Article 50(2) only.

The EU rules can reach companies outside the EU, because Article 2(1) covers providers placing AI systems on the EU market wherever they are established. Article 99(4) puts Article 50 in the fine tier capped at EUR 15,000,000 or 3% of worldwide annual turnover, whichever is higher, with separate treatment for small and medium-sized companies. Under Article 99(6a), added by the Omnibus, a small mid-cap faces the lower of the two figures. These are ceilings, not set amounts.

Content Provenance and Watermark Laws: SB 942, SB 1000, Washington and Utah

Provenance laws do not usually tell a person anything directly. They require data embedded in a file, so that a tool can later show where the file came from.

California AI Transparency Act After SB 1000

Many guides still describe the version before SB 1000. As rewritten:

• Threshold. A "covered provider" no longer needs 1,000,000 users. The test is whether the person creates, codes or otherwise produces a generative AI system publicly accessible in California.
• Tool. The "AI detection tool" is now a "disclosure verification tool." It must be offered at no cost, and a provider may instead direct users to a compliant third-party tool.
• Manifest option. The duty to offer users a manifest disclosure option is deleted.
• Latent disclosure. It conveys the provider's name, the system's name and version, the time and date, a unique identifier, and whether the system created or altered the content.
• Licensees. A provider that knows a licensee modified a system out of compliance has 72 hours to end the authorization or notify the licensee. The licensee then has 96 hours to report its fix.

The chapter does not apply to products that provide exclusively non-user-generated videogames. Before January 1, 2029, it does not apply to a system designed to primarily function as assistive technology, and falsely claiming that status carries $50,000 per violation under section 22757.4.1 until that section's repeal on that date.

From January 1, 2027, large online platforms (more than 2,000,000 unique monthly users in the preceding 12 months) must detect compliant provenance data on content they distribute and give users a way to see it. From the same date, a generative AI hosting platform "shall not knowingly make available a GenAI system that does not place disclosures" as the act requires. Makers of capture devices first produced for sale in California on or after January 1, 2028 must offer a latent disclosure option and embed latent disclosures by default, to the extent technically feasible and compliant with widely adopted specifications.

Washington E2SHB 1170, Connecticut and Utah

Washington's law is often called an AI disclosure law, but it does not make private companies tell each user that content came from AI. Its private-sector duty is to embed provenance data. It does not apply to business-to-business uses, products that provide exclusively video game or interactive experiences, or systems used solely for upscaling, noise reduction or compression. Minor changes such as brightness, resizing, cropping and format conversion do not count as material alterations.

Connecticut's section uses the same structure. A provider must also use reasonable methods, including the Coalition for Content Provenance and Authenticity standard, to make the data hard to remove.

Utah's HB 276 adds duties for platforms that exceeded 2,000,000 unique monthly users in the preceding 12 months, which must detect compliant provenance data, disclose its availability and let users inspect it. Capture device makers must add a latent disclosure on devices produced for sale in Utah on or after January 1, 2028, though they may let users turn it off.

Telling People They Are Dealing With AI: Utah SB 226 and Maine LD 1727

These three laws are the closest to what many readers mean by an "AI disclosure law," and all three are narrow.

Utah SB 226. Under Utah Code 13-77-103(1), a supplier using generative AI to interact with an individual in connection with a consumer transaction must disclose that the person is interacting with generative AI "and not a human, if the individual asks or otherwise prompts the supplier about whether artificial intelligence is being used." The prompt must be "a clear and unambiguous request."

Section 13-77-103(2) reaches individuals providing services in a regulated occupation, meaning one the Department of Commerce regulates and that requires a license or state certification. They must "prominently disclose" generative AI in a high-risk AI interaction, verbally at the start of a verbal interaction and in writing before a written one. A high-risk interaction involves collecting sensitive personal information, or personalized advice that could reasonably be relied on for significant personal decisions, or other uses defined by rule.

The chapter's definition of generative AI is narrower than the everyday term. A system must be trained on data, be designed to simulate human conversation with a consumer, and generate "non-scripted outputs similar to outputs created by a human, with limited or no human oversight." Section 13-77-104 gives a safe harbor where the system clearly and conspicuously discloses what it is at the outset and throughout the interaction.

Maine LD 1727. Under 10 MRSA 1500-DD:

A person may not use an artificial intelligence chatbot or any other computer technology to engage in trade and commerce with a consumer in a manner that may mislead or deceive a reasonable consumer into believing that the consumer is engaging with a human being unless the consumer is notified in a clear and conspicuous manner that the consumer is not engaging with a human being.

The trigger is the risk of deception, so a chatbot that could not mislead a reasonable consumer does not trip the section by being a chatbot. The rule also reaches "any other computer technology."

California's bot law. Business and Professions Code 17941, operative since July 1, 2019, makes it unlawful to use a bot to communicate with a person in California online "with the intent to mislead the other person about its artificial identity" in order to knowingly deceive them about the content of the communication, to encourage a sale or influence a vote. A person who discloses that it is a bot is not liable under the section. A bot is an automated online account where all or substantially all of the actions or posts are not the result of a person.

Companion and conversational AI laws in other states have their own not-human notices, and many start in 2027. Those laws are outside this count, and our AI companion chatbot laws page covers them.

AI in Advertising: New York GBL 396-b, California SB 1050 and Pennsylvania Act 99

Three general laws reach AI in ads, apart from the real estate and election rules below. Each has a different trigger, so an ad that needs a disclosure in one state may not need one in another.

New York GBL 396-b(3). A person in the business of dealing in property or services who produces or creates an ad "shall conspicuously disclose in such advertisement that a synthetic performer is in such advertisement, where such person has actual knowledge." A synthetic performer is a digitally created asset intended to create the impression of a human performer "who is not recognizable as any identifiable natural performer." So the rule targets invented people, not likenesses of real ones.

The section exempts ads for expressive works where the synthetic performer's use is consistent with the work, audio ads, and AI used solely to translate a human performer's language. It does not apply to the media that carry the ad.

California SB 1050. From January 1, 2027, it is unlawful to create and cause to be published an ad that prominently includes a synthetic performer without a clear and conspicuous disclosure. An advertising medium is barred only after a court order is served on it.

Pennsylvania Act 99 of 2026. Governor Shapiro approved SB 806 on October 22, 2026. It says it "shall take effect in 60 days," and our count from approval gives about December 21, 2026. Section 4(a) of the act says:

An advertiser may not knowingly, recklessly or willfully create and disseminate or cause to be created and disseminated an advertisement that uses synthetic content in a manner that creates a false consumer perception about the nature, origin or authenticity of the goods or services being advertised, without clear and conspicuous disclosure.

So an ad that uses AI imagery without creating that false perception falls outside the sentence. Section 4(b) excepts evidently fictional content, parody, satire, commentary or criticism, and content "limited to text." A disclosure "shall not be construed to limit any recourse under law."

Other Disclosure Rules: Pricing, Training Data and Subscriptions

New York GBL 349-a. An entity that sets a price using "personalized algorithmic pricing" and shows it to a New York consumer must include the statement "THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA". The section exempts entities subject to the insurance law, certain financial institutions and a price below an existing subscription price.

In NRF v. James, the district court dismissed a First Amendment challenge with prejudice on October 8, 2025. The Second Circuit's record lists the appeal, No. 25-2818, as argued October 19, 2026, and its decisions page showed no decision when searched on November 4, 2026.

California AB 2013. Developers of generative AI systems must publish training-data documentation. A court denied a preliminary injunction in X.AI v. Bonta on March 4, 2026, and the Ninth Circuit's calendar, read November 4, 2026, lists argument in No. 26-1591 for November 18, 2026.

Connecticut subscriptions. PA 26-100 sec. 46 bars a subscription-based provider from starting or renewing a subscription, or collecting a fee for one, unless it gave written notice of the key terms and the consumer gave written notice of accepting them. It covers providers with more than one million users per month.

Laws Outside the Count

These laws also require AI-related disclosure, but none of them counts toward the seven. We keep them apart so that the count measures one thing.

Synthetic Media in Elections and Political Ads

Many states require disclaimers on, or limit, deceptive AI media in election communications. They differ on who is covered and what the disclaimer must say, and our deepfake laws guide covers them.

The wording can be exact. Wisconsin's Act 123 requires audio to say "Contains content generated by AI" at the beginning and end. Louisiana's Act 559 requires a disclosure at the start of a campaign call created using AI with a public figure's voice.

Courts have limited several of these laws. A federal court permanently enjoined California's AB 2839 as to the named plaintiffs, and a separate judgment held AB 2655 preempted by Section 230 as applied to X Corp., Rumble and any such provider of an interactive computer service; the state has appealed. Hawaii's state defendants are permanently enjoined from enforcing Act 191. A September 16, 2026 clerk's docket entry records a preliminary injunction against Montana's SB 25 limited to the named plaintiffs; we did not read the order.

Real Estate Ads: Wisconsin 2025 Act 69

Wis. Stat. 452.136(1m), added by Act 69, reads: "A licensee shall in all advertising disclose if the advertising has been altered or modified using technology, including artificial intelligence, to add, remove, or change elements of the property, that creates a false or misleading impression of the property."

It covers licensees under the real estate chapter, not advertisers in general, and it reaches alterations that create a false or misleading impression, not every use of AI. Section 9 of the act says it takes effect January 1, 2027.

What Changes in AI Disclosure Law in 2027

Bills, not laws. As of November 4, 2026, the bills we saw were still bills. The District of Columbia's B26-0858, introduced October 9, 2026, would require disclosure when a real property ad uses an altered image. Minnesota's SF 1886, on disclosure when a person communicates with AI, was amended and re-referred on March 18, 2026, per the bill page. Delaware's HB 306, on chatbot disclosure, did not pass, and its bill page says the General Assembly has ended.

We make no prediction about any bill, and our AI laws tracker is updated in place.

Related AI Law Pages

• AI laws by state
• AI companion chatbot laws
• Deepfake laws by state
• California AI laws
• Connecticut AI laws
• Maine AI laws
• New York AI laws
• Pennsylvania AI laws
• Utah AI laws
• Washington AI laws

Frequently Asked Questions

Is there a federal law requiring AI labels?

As of November 4, 2026, we found no federal statute in the public laws of the 119th Congress that requires labels on AI content. The FCC's February 2024 ruling treats AI-generated voices in calls as artificial under the Telephone Consumer Protection Act, but that is a ruling on an existing statute, not an AI label.

Which states have AI disclosure laws?

By our count of enacted laws as of November 4, 2026, and subject to the search limits on our AI laws tracker, at least seven states have general AI disclosure or labeling laws, meaning laws not limited to elections, to one sector or to chatbot safety: California, Connecticut, Maine, New York, Pennsylvania, Utah and Washington. Pennsylvania's law takes effect about December 21, 2026, and Washington's takes effect February 1, 2027.

Do I have to label AI-generated content?

It depends on what you do and where. None of the laws on this page is a general rule that all AI content must carry a visible label. They put specific duties on makers of large or public generative AI systems, on advertisers in certain situations, on businesses whose chatbots may mislead consumers, and on some licensed professionals. Election and health care laws add their own rules.

What AI video rules took effect in August 2026?

On August 2, 2026, California's AI Transparency Act became operative and the EU AI Act's Article 50 transparency duties began to apply. California's act requires covered providers to embed latent disclosures in image, video and audio content. In the EU, providers of systems already on the market before that date have until December 2, 2026 to comply with Article 50(2) marking.

Is the California AI Transparency Act in effect?

Yes. SB 942 became operative August 2, 2026, and SB 1000 rewrote it with immediate effect on September 30, 2026. Large online platform and hosting platform duties start January 1, 2027, and capture device duties start January 1, 2028.

Does SB 942 still have a 1,000,000-user threshold?

No. SB 1000 removed it on September 30, 2026. A covered provider is now a person that creates, codes or otherwise produces a generative AI system publicly accessible in California. Connecticut, Utah and Washington still use one-million user thresholds in their own laws.

What is a latent disclosure?

It is information embedded in a file, not a label a viewer sees. Under California's act it conveys the provider's name, the system's name and version, the time and date, a unique identifier, and whether the system created or altered the content.

What does EU AI Act Article 50 require?

Among other things, providers of systems generating synthetic audio, image, video or text must mark outputs in a machine-readable format, subject to exceptions, and deployers must disclose deepfakes, with a lighter rule for evidently artistic or satirical work. The duties have applied since August 2, 2026.

Do businesses have to tell customers they are using a chatbot?

Not under one national rule. In Maine, a clear and conspicuous notice is needed when a chatbot or other computer technology may mislead a reasonable consumer into thinking it is human. In Utah, a supplier must disclose generative AI when a person clearly asks. California bars using a bot online to knowingly deceive a person about its artificial identity to encourage a sale, unless it discloses that it is a bot. Other states have separate rules for companion and conversational AI chatbots, many starting in 2027.

Do AI-generated ads need a label?

Sometimes. New York requires disclosure of a synthetic performer where the person producing the ad has actual knowledge. California's SB 1050 adds its own synthetic performer rule from January 1, 2027. Pennsylvania's Act 99, from about December 21, 2026, applies when synthetic content creates a false consumer perception about goods or services.

Does Pennsylvania require labels on AI ads?

Not on every AI ad. Act 99 applies when an advertiser knowingly, recklessly or willfully uses synthetic content in a way that creates a false consumer perception about the nature, origin or authenticity of the goods or services, without clear and conspicuous disclosure. It excepts evidently fictional content, parody, satire, commentary or criticism, and content limited to text.

Do AI-edited real estate photos have to be disclosed?

In Wisconsin, starting January 1, 2027, a real estate licensee must disclose advertising altered with technology, including AI, to add, remove or change elements of the property in a way that creates a false or misleading impression. A District of Columbia bill on altered listing photos was introduced October 9, 2026, and is not law.

Do AI political ads need a disclaimer?

In many states, some do, under election laws that vary widely. Several have been limited in court. They are outside our seven-state count, and our deepfake laws guide covers them.