# AI in Political Ads: State Disclosure Laws, Election Deepfake Bans, Court Rulings, and Federal Rules

Source: https://superintelligencecompliance.ai/guides/ai-political-ads  
Publisher: ZSKFL Management  
Last updated: 2026-11-16. Not legal advice.

> No federal law requires AI labels on political ads as of November 16, 2026. States set label text and pre-election deepfake bans; courts enjoined some.

## At a glance

- As of November 16, 2026, no federal statute or FEC or FCC rule requires political ads to disclose AI. The duty comes from state law, and each state sets its own trigger, label text, and exceptions.
- As of October 13, 2026, NCSL counted 31 states that have enacted laws regulating the use of deepfakes in political messaging. Some require a fixed label on AI content in paid political ads. Others ban distributing a deceptive deepfake, usually of a candidate, inside a pre-election window (30 to 90 days in most of the laws below), with knowledge and intent elements that vary by state. Most of those exempt content that carries the required disclaimer.
- Courts have blocked some of these laws, mostly for named plaintiffs only. California’s AB 2839 is permanently enjoined as to the named plaintiffs (rewritten by AB 502 in September 2026) and AB 2655 as to X Corp. and Rumble (the state appealed the AB 2655 judgment), Hawaii’s Act 191 is permanently enjoined, and in September 2026 federal courts in Montana and Vermont issued preliminary injunctions limited to the plaintiffs before them (in Vermont, to one investigative demand and one video).
- Google and Meta require advertisers to disclose certain synthetic or AI-edited content in election and political ads under their own policies. Those are voluntary platform rules, not law.

## Do political ads have to disclose AI?

Only where a state law says so, and only for the ads that law covers. As of November 16, 2026, no federal statute requires an AI label on political ads, the FEC has declined to open an AI rulemaking, and we found no adoption of the FCC’s 2024 proposal for broadcast disclosures. State laws fill the gap in two ways: a label requirement for AI content in political ads, or a ban on distributing a deceptive deepfake, usually of a candidate, inside a window before an election. Most of the bans treat a disclaimer as a safe harbor.

## Which states require a label on AI-generated political ads?

Except in Washington, which reaches electioneering communications as Title 29B RCW defines them, these laws apply whenever the trigger is met, with no pre-election window. The triggers differ. California, Michigan, Wisconsin, and Utah attach the label to AI or synthetic content in a covered ad. Florida adds a depiction and intent test, and New York reaches only materially deceptive media distributed with actual knowledge.

**State AI disclosure laws for political ads, examples as of November 16, 2026**

| State | Law | Disclosure window | Label required | Exceptions | Enforcement | Court status |
|---|---|---|---|---|---|---|
| California | Government Code 84514 (AB 2355, Chapter 260, Statutes of 2024), effective January 1, 2025 | None; applies to a qualified political advertisement a committee creates, originally publishes, or originally distributes | “Ad generated or substantially altered using artificial intelligence.” in the manner the disclaimer rules in sections 84504 to 84504.5 prescribe | Media immaterially altered by AI, such as a cosmetic adjustment, color edit, cropped image, or resized image, is not covered; Section 230 rights are not altered | Fair Political Practices Commission injunctive relief or administrative or civil remedies; a violation is not a misdemeanor | Not part of the *Kohls v. Bonta* judgments |
| Michigan | MCL 169.259 (2023 PA 263), effective February 13, 2024 | None; any paid advertisement relating to a Michigan candidate, election, or ballot question that contains an image, audio, or video generated in whole or substantially with AI | A clear and conspicuous statement that the ad “was generated in whole or substantially by artificial intelligence,” with size, duration, and language rules by format | Bona fide news coverage that acknowledges the content is not accurate; stations paid to broadcast the ad; news distribution platforms with a written disclosure policy given to advertisers; satire or parody; businesses regulated by the Michigan Public Service Commission or the FCC | State civil infraction: up to $250 for a first violation and up to $1,000 for each later one; each ad distributed or aired is a separate violation | None found in the sources we checked |
| Wisconsin | Wis. Stat. 11.1303(2m) (2023 Wis. Act 123) | None; audio and video communications covered by 11.1303(2)(a) or (b) that contain express advocacy, issue advocacy, or support for or opposition to a referendum, and contain synthetic media | Audio: “Contains content generated by AI.” at the beginning and end. Video: “This video content generated by AI,” “This audio content generated by AI,” or “This content generated by AI,” throughout each portion containing synthetic media | Broadcasters and other hosts or carriers, unless the person is a committee responsible for the communication; Section 230 rights are not altered; the Ethics Commission may set limited exceptions by rule | Forfeiture of up to $1,000 for each intentional violation | None found in the sources we checked |
| Florida | Fla. Stat. 106.145 (Chapter 2024-126) | None; political ads, electioneering communications, and other political ads with generative AI content that appears to depict a real person performing an action that did not occur, created with intent to injure a candidate or to deceive regarding a ballot issue | “Created in whole or in part with the use of generative artificial intelligence (AI).” with font, height, and audio rules | None stated in the section | First-degree misdemeanor for a person identified as paying for, sponsoring, or approving the ad; any person may complain to the Florida Elections Commission, which must provide an expedited hearing | None found in the sources we checked |
| Utah | Utah Code 20A-11-1104 (Chapter 172, 2024 General Session), effective May 1, 2024 | None; paid audio or visual communications by a candidate committee, political action committee, political issues committee, political party, or a person using a contribution, intended to influence voting in Utah, that contain synthetic media | “Contains content generated by AI.” at the beginning and end of audio; for visual media, one of four fixed statements throughout each portion; online communications must also carry embedded tamper-evident digital content provenance | A person that solely provides the technology used is not a “creator” | Civil penalty of up to $1,000 for each violation in a civil action | None found in the sources we checked |
| New York | Election Law 14-106(5), added by Chapter 58 of 2024 and amended by Chapter 169 of 2024 | None; a political communication produced by or including materially deceptive media, distributed with actual knowledge that it is materially deceptive | “This (image, video, or audio) has been manipulated” | Satire or parody; qualifying news reporting; broadcasters and streaming services with consistent disclaimer requirements; initial dissemination by a platform of content provided by another | A depicted candidate may seek injunctive relief, court costs, and attorneys’ fees, proving the use by clear and convincing evidence | None found in the sources we checked |
| Washington | RCW 29B.35.010 to 29B.35.030 (2023 c 360; recodified from chapter 42.62 RCW effective January 1, 2026) | Electioneering communications as Title 29B RCW defines them: communications that identify a state, local, or judicial candidate, are published within 60 days before that election, and cost $1,000 or more, alone or with the same sponsor’s other such communications (RCW 29B.10.220) | No label is mandated; “This (image/video/audio) has been manipulated,” displayed as the statute prescribes, is an affirmative defense | The sponsor, not the disseminating medium, is liable unless the medium removes the disclosure or alters the content (not for broadcasters transmitting under 47 U.S.C. 315); interactive computer services are not treated as publishers of others’ content | A depicted candidate may sue for injunctive relief and damages against the sponsor, proving synthetic media by clear and convincing evidence | None found in the sources we checked |

## Which states ban deceptive election deepfakes, and when?

These laws reach narrower content than the label laws: a deceptive depiction, usually of a candidate, made or shared with the knowledge or intent each statute sets out. Most reach anyone who distributes it inside the window, not only advertisers; Arizona, Montana, and Texas are narrower (see the liability section).

**State election deepfake bans, examples as of November 16, 2026**

| State | Law | Disclosure window | Label required | Exceptions | Enforcement | Court status |
|---|---|---|---|---|---|---|
| California | Elections Code 20010 (AB 730, 2019, as amended by AB 972, 2022) | Within 60 days of an election at which the candidate will appear on the ballot; distribution with actual malice and intent to injure the candidate’s reputation or to deceive a voter | “This _____ has been manipulated.” (image, video, or audio) as a safe harbor | Bona fide news broadcasts that acknowledge questions about authenticity; stations paid to broadcast; news websites and periodicals that clearly state the media is not accurate; satire or parody; Section 230 rights not altered | Candidate injunctive relief and damages, proving the violation by clear and convincing evidence | This version is repealed on January 1, 2027 by its own terms, when a later version on superimposed candidate photographs becomes operative |
| California | Elections Code 20012 (AB 2839, 2024), rewritten by AB 502 (Chapter 333, Statutes of 2026), an urgency statute chaptered September 19, 2026 | 30 days before an election for depictions of candidates and elected officials, and from 30 days before through 38 days after it for elections officials and voting equipment; knowing distribution, with malice (knowing the content is false, or with reckless disregard for the truth), of an advertisement or other election communication containing materially deceptive content | None for most content; a candidate portraying themselves is exempt with the Government Code 84514 label or “This ____ has been manipulated.” in the prescribed type and placement | Bona fide news broadcasts and news publications that state the content is not accurate; paid broadcasts where the station gave buyers consistent requirements, or where federal law requires it to air the ad or bars it from altering the message; satire or parody a reasonable person would understand, or labeled as such; interactive computer services are not liable | Injunctive relief and damages sought by a depicted individual, a candidate or committee in the election, or an elections official, proving the violation by clear and convincing evidence, with attorneys’ fees; no damages against a broadcasting station or website that did not create the content | The August 29, 2025 judgment permanently enjoined AB 2839 as enacted, as to the named plaintiffs, and was not appealed; we found no court order on the rewritten section as of November 16, 2026 |
| California | AB 2655 (2024) | Not summarized here; the law is enjoined as shown | Permanently enjoined as to X Corp. and Rumble (August 20, 2025), on appeal to the Ninth Circuit; the state stipulated not to enforce it against any interactive computer service provider unless an appellate court vacates the judgment |
| Michigan | MCL 168.932f (2023 PA 265), effective February 13, 2024 | Within 90 days before an election; the person must know the media falsely represents the depicted individual and intend to harm a candidate and change voting behavior by deception, with the distribution reasonably likely to cause both results | A disclaimer that the media “has been manipulated by technical means and depicts speech or conduct that did not occur” (sample text sufficient, not necessary), plus a citation to the original source when existing media was edited | The disclaimer safe harbor is the only exception the section states | Misdemeanor (up to 90 days, $500, or both); felony within 5 years of a prior conviction; permanent injunctions sought by the attorney general, the depicted individual, a candidate, or a voter organization; no preliminary injunctive relief | None found in the sources we checked |
| Minnesota | Minn. Stat. 609.771 | Within 90 days before a party nominating convention, or after absentee voting starts for a presidential nomination primary or a regular or special state or local primary or general election; dissemination without the depicted individual’s consent, knowing or with reckless disregard about whether it is a deepfake, with intent to injure a candidate or influence an election | None; the statute provides no disclaimer defense | A broadcaster or cable system whose dissemination of a candidate-produced deepfake is required by federal law | Crime, with higher penalties for repeat or violent-intent offences; injunctions sought by the attorney general, a county or city attorney, the depicted individual, or a candidate; forfeiture of the nomination or office for a convicted candidate for state or local office | *Kohls v. Ellison*: Eighth Circuit affirmed the denial of a preliminary injunction (February 9, 2026) without deciding constitutionality; rehearing denied March 31, 2026 |
| Texas | Election Code 255.004(d) (SB 751, 2019) | Publication or distribution within 30 days of an election, by a person who created the deep fake video with intent to injure a candidate or influence the result | None; the subsection has no disclaimer defense | None stated in subsections (d) and (e); covers video only | Class A misdemeanor (Election Code 255.004(c)) | None found in the sources we checked |
| Arizona | A.R.S. 16-1024 | Within 90 days before an election at which the candidate appears on the ballot; a creator who knows the media is a deceptive and fraudulent deepfake | A clear and conspicuous disclosure that conveys to a reasonable person that the media includes AI-generated content (no fixed wording) | Satire or parody; interactive computer services; providers of the technology are not creators | The civil penalty in A.R.S. 16-937(B) for each day of distribution without the disclosure | None found in the sources we checked |
| Vermont | 17 V.S.A. 2031 to 2033 (Act 75, S.23, signed March 5, 2026, effective on passage) | Within 90 days before an election in Vermont; synthetic media the person knows is deceptive and fraudulent | “This media has been manipulated or generated by digital technology and depicts speech or conduct that did not occur.” | Qualifying news broadcasts; paid broadcasts; ads federal law requires a broadcaster to carry; news publications that state the media is not accurate; satire or parody; telecommunications, information service, and interactive computer service providers for others’ content | Fine of up to $1,000 for a knowing and intentional violation, or up to $5,000, $10,000, or $15,000 where intent to cause violence or bodily harm, a prior violation within 5 years, or both apply; candidate injunctive relief; attorney general or State’s Attorney actions and civil investigative demands | *Poitras v. Clark* (D. Vt.): preliminary injunction (September 4, 2026) limited to one investigative demand and one video; the court did not decide the facial challenge |
| Montana | MCA 13-35-801 to 13-35-804 (SB 25, 2025) | Within 60 days of the initiation of voting; a person working in an official election capacity who pays for or sponsors a communication it knew or should have known is a deepfake of a candidate or party | “This _____ (image/audio/video/multimedia) has been significantly edited by artificial intelligence and depicts speech or conduct that falsely appears to be authentic or truthful” | Qualifying news broadcasts; paid broadcasts or ads federal law requires; satire or parody; distribution platforms that gave the purchaser consistent disclaimer requirements; interactive computer, information, and telecommunications service providers for others’ content | Candidate or party actions for an injunction, costs, and fees, plus actual damages and up to $10,000 in punitive damages; complaints to the Commissioner of Political Practices, with civil penalties on a first sufficiency finding, misdemeanor referral on a second, and felony referral on a third | *Accountability in State Government v. Knudsen* (D. Mont.): preliminary injunction (September 16, 2026) as to the named plaintiffs only |
| Hawaii | Act 191 (2024) | Not summarized here; the law is enjoined | *The Babylon Bee, LLC v. Lopez* (D. Haw.): permanently enjoined January 30, 2026 |

## Have courts blocked state AI election laws?

Several, but most orders protect only the parties that sued, and each ruling below reaches one statute.

- **California.** In *Kohls v. Bonta*, No. 2:24-cv-02527 (E.D. Cal.), a final judgment on August 20, 2025 declared that AB 2655 violates and is preempted by Section 230 as applied to X Corp. and Rumble and permanently enjoined its enforcement against them. The state also stipulated not to enforce AB 2655 against any provider of an interactive computer service; that agreement ends if an appellate court vacates the AB 2655 judgment on the state’s appeal. On August 29, 2025 the court held that AB 2839 fails strict scrutiny in its entirety and permanently enjoined its enforcement against the named plaintiffs. The state appealed the AB 2655 judgment to the Ninth Circuit on September 19, 2025 (No. 25-6138) and did not appeal the AB 2839 judgment; we found no appellate decision as of November 16, 2026. AB 502, an urgency statute chaptered September 19, 2026, rewrote the AB 2839 section (Elections Code 20012); we found no order on the new text.
- **Hawaii.** On January 30, 2026, in *The Babylon Bee, LLC v. Lopez*, No. 1:25-cv-00234 (D. Haw.), a federal court held that Act 191 violates the First Amendment and is unconstitutionally vague, and permanently enjoined the state defendants from enforcing it.
- **Minnesota.** In *Kohls v. Ellison*, No. 25-1300, the Eighth Circuit on February 9, 2026 affirmed the denial of a preliminary injunction against 609.771, holding only that the district court did not abuse its discretion. It did not decide whether the law is constitutional.
- **Montana.** In *Accountability in State Government v. Knudsen*, No. 6:26-cv-00038 (D. Mont.), the court on September 16, 2026 found the plaintiffs likely to succeed in showing the Act imposes an unlawful content- and viewpoint-based restriction, and enjoined the attorney general, a county attorney, and the Commissioner of Political Practices from enforcing it against the named plaintiffs. Citing *Trump v. CASA*, it declined to enjoin enforcement against anyone else. The ruling is preliminary, and the case continues.
- **Vermont.** In *Poitras v. Clark*, No. 2:26-cv-00234 (D. Vt.), the court on September 4, 2026 restrained the attorney general’s office from enforcing one civil investigative demand and from any Act 75 enforcement against the plaintiff over one video. It wrote that it was “not considering whether Act 75 is unconstitutional.” As of its last entry on October 26, 2026, the docket showed a motion to dismiss fully briefed and no ruling on it.
Outside those orders, as the [federal preemption guide](https://superintelligencecompliance.ai/guides/federal-ai-preemption) explains, state laws still apply unless and until a court, or an act of Congress, says otherwise.

## What does federal law require for AI in political ads?

No AI disclosure. Federal rules reach fraudulent misrepresentation of campaign authority and AI-voice robocalls, and we found no adoption of the FCC’s 2024 proposal for broadcast disclosures.

- **FEC.** Public Citizen petitioned on July 13, 2023 for a rule applying the ban on fraudulent misrepresentation of campaign authority to deliberately deceptive AI campaign ads. On September 26, 2024 the FEC published its decision not to open a rulemaking and an interpretive rule, effective that day, that 52 U.S.C. 30124 and 11 CFR 110.16 apply irrespective of the technology used. The FEC says the interpretive rule does not bind the Commission or any member of the public. The statute reaches a candidate or agent who fraudulently misrepresents acting for another candidate or party on a damaging matter, and any person who does so to solicit contributions. It is not a disclosure rule.
- **FCC proposal.** In a notice of proposed rulemaking published August 5, 2024 (MB Docket No. 24-211, FCC 24-74), the FCC proposed that radio and television broadcast stations; cable operators, Direct Broadcast Satellite providers, and Satellite Digital Audio Radio Service licensees engaged in origination programming; and section 325(c) permit holders give an on-air announcement for political ads (candidate and issue ads) that contain AI-generated content, and note it in their online political files. As of November 16, 2026, we found no later Federal Register document adopting, withdrawing, or terminating it, and the FCC’s electronic filing system listed the docket without a closing entry.
- **Robocalls.** The FCC’s Declaratory Ruling FCC 24-17, adopted February 2, 2024, confirms that the TCPA’s restrictions on an “artificial or prerecorded voice” encompass current AI technologies that resemble human voices or generate call content using a prerecorded voice, so those calls need the called party’s prior express consent absent an emergency purpose or exemption. Where such a call introduces an advertisement or contains telemarketing, the FCC’s rules require prior express written consent (47 CFR 64.1200(a)(2) and (3), cited in footnote 13 of the ruling). It is not limited to political calls.
- **Robocall enforcement.** In a forfeiture order adopted September 26, 2024 (FCC 24-104), the FCC imposed a $6,000,000 penalty on Steve Kramer for robocalls carrying a deepfake AI voice message that imitated President Biden, sent to potential New Hampshire voters two days before the state’s 2024 Democratic presidential primary, with misleading caller ID in violation of the Truth in Caller ID Act.

## Which federal AI election bills are pending?

None has passed either chamber. In govinfo bill status records checked November 16, 2026, each of these 119th Congress bills had been referred to committee and had no later action:

- **Protect Elections from Deceptive AI Act**: S. 1213 (Senate, introduced March 31, 2025, Rules and Administration) and H.R. 5272 (House, September 10, 2025, House Administration). It would bar knowingly distributing materially deceptive AI-generated audio or visual media of a federal candidate to influence an election or solicit funds, with news and satire exceptions.
- **AI Transparency in Elections Act of 2026**: H.R. 9922 (House, July 23, 2026, House Administration). It would require a clear and conspicuous statement in covered communications financed by a disbursement that contain an image, audio, or video substantially generated by AI.
- **AI Ads Act**: S. 5135 (Senate, July 27, 2026, Rules and Administration) and H.R. 9985 (House, July 30, 2026, House Administration).
- **FAIR Elections Act of 2026**: S. 4774 (Senate, June 11, 2026, Rules and Administration).
- **REAL Political Advertisements Act**: H.R. 10814 (House, October 15, 2026, House Administration).
- **Securing Elections From AI Deception Act**: H.R. 10841 (House, October 19, 2026, Energy and Commerce).
The NO FAKES Act of 2026 (S. 4591) is a likeness bill, but its preemption clause speaks to election laws. As reported, it would preempt state causes of action for voice and visual likeness rights in a digital replica in an expressive work, except, among others, causes of action under state statutes specifically on sexually explicit or election-related digital replicas (section 2(g)). It was placed on the Senate calendar on June 24, 2026 and, as of November 16, 2026, had not passed either chamber.

## What do Google and Meta require?

Disclosure of certain AI content under their own ad policies. These are voluntary platform rules, not law.

- **Google.** Its political content policy says advertisers must disclose election ads that contain synthetic or digitally altered content that depicts real or realistic-looking people or events inauthentically, by selecting a checkbox in campaign settings. Google generates an in-ad disclosure for some formats; for others the advertiser must add a prominent disclosure. Edits inconsequential to the claims made in the ad are out of scope.
- **Meta.** Its standards for ads about social issues, elections, or politics require advertisers to disclose when an ad contains a photorealistic image or video, or realistic-sounding audio, created or edited with third-party generative AI tools in the cases the policy lists. Meta says it rejects ads that should have been disclosed and that repeated failure may lead to penalties. It also says that from June 1, 2026 it uses automated detection to identify ad media created or edited with third-party generative AI tools and add an “AI Info” label.

## Who is liable: the campaign, the agency, the broadcaster, or the platform?

Usually the person who pays for, sponsors, or creates the ad, and under most window bans, anyone who distributes it with the required knowledge and intent.

- **Committees, sponsors, and payers** carry the label duty under California Government Code 84514 (committees), Utah (committees, parties, and persons using a contribution, with civil actions against the creator or sponsor), and Florida (the person named as paying for, sponsoring, or approving the ad). Washington makes the sponsor liable, and Montana’s ban reaches a person working in an official election capacity who pays for or sponsors the communication.
- **Creators** carry the duty in Arizona and Texas, whose statutes reach the person who created the media. Michigan’s label law binds any person or entity that creates, originally publishes, or originally distributes the ad.
- **Distributors** who meet the knowledge and intent elements are covered by California Elections Code 20010 and 20012, Michigan’s deepfake law (MCL 168.932f), New York, Vermont, and Minnesota.
- **Broadcasters** paid to air an ad are exempt under California Elections Code 20010, Michigan’s label law (MCL 169.259), and the Vermont and Montana laws, and Wisconsin exempts hosts and carriers that are not the responsible committee. The exemption is narrower elsewhere. New York’s covers broadcasters and streaming services with consistent disclaimer requirements, Minnesota’s covers only dissemination of a candidate-produced deepfake that federal law requires, and Michigan’s deepfake law and Texas’s subsections (d) and (e) state none.
- **Platforms** keep their Section 230 protection under several statutes (California, New York, Wisconsin, Washington, Arizona, Vermont, Montana), and a court held California’s AB 2655 preempted by Section 230 as applied to X Corp. and Rumble.
An ad agency that produces an ad can be a “creator” or distributor under some of these laws even when the campaign is the sponsor. Utah and Arizona exclude a person who only provides the technology from “creator”.

## What should campaigns, agencies, broadcasters, and platforms do?

- Map every state where an ad will run against both kinds of law: a label law triggered by AI or synthetic content in the ad (Florida and New York add intent or knowledge tests), and a window law that bans deceptive deepfakes, usually of candidates, before the election unless a safe harbor applies
- Use each state’s exact label text and format rules; California, New York, Wisconsin, Utah, Florida, Montana, and Vermont fix the wording, Michigan’s deepfake law gives sample text that is sufficient but not required, and Arizona requires only a clear and conspicuous disclosure; in Washington and under the window bans, the statutory text is a defense or safe harbor, not a mandate
- For paid online communications covered by Utah’s law, embed tamper-evident digital content provenance that names the initial author, later editors, and any generative AI use
- Keep records of every AI tool used on an ad, the edits made, and the source files, so you can show whether an edit was immaterial or substantial
- Do not rely on a disclaimer in Minnesota or Texas, where the statutes provide none, or under California Elections Code 20012 except for a candidate’s self-portrayal or labeled satire
- Broadcasters: keep the paid-ad and bona fide news exceptions documented, and in New York give advertisers your disclaimer requirements, which the exception requires; in Minnesota the exception covers only dissemination of a candidate-produced deepfake that federal law requires
- Platforms: keep a written AI disclosure policy and give it to political advertisers where a state ties the exemption to one, as Michigan’s label law does for news distribution platforms
- Advertisers on Google or Meta: make the AI disclosure each platform’s policy requires for each ad (on Google, a checkbox in campaign settings)
- Get prior express consent before any TCPA-covered call that uses an AI-generated voice, absent an emergency purpose or exemption, and prior express written consent where the call introduces an advertisement or contains telemarketing; do not use misleading caller ID
- Track the court orders by party: an injunction for named plaintiffs does not protect other speakers
How this fits the other guides: this page sits under the [US state AI laws guide](https://superintelligencecompliance.ai/guides/us-state-ai-laws) and expands the election section of the [deepfake laws guide](https://superintelligencecompliance.ai/guides/deepfake-laws). The [New York](https://superintelligencecompliance.ai/guides/new-york-ai-laws), [California](https://superintelligencecompliance.ai/guides/california-ai-laws), and [Texas](https://superintelligencecompliance.ai/guides/texas-ai-law-hb-149) guides cover those states’ other AI laws.

This guide is general information as of November 16, 2026, not legal advice; confirm requirements with counsel or the election regulator before acting.

## Sources

- [Deepfakes in Elections and Campaigns](https://www.ncsl.org/elections-and-campaigns/artificial-intelligence-ai-in-elections-and-campaigns), National Conference of State Legislatures, updated October 13, 2026.
- [California Government Code 84514](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=GOV§ionNum=84514), California Legislative Information, accessed November 16, 2026.
- [California Elections Code 20010 (both versions)](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=ELEC§ionNum=20010), California Legislative Information, accessed November 16, 2026.
- [AB 502, Elections: deceptive media in advertisements (Chapter 333, Statutes of 2026)](https://leginfo.legislature.ca.gov/faces/billNavClient.xhtml?bill_id=202520260AB502), California Legislative Information, chaptered September 19, 2026.
- [MCL 169.259](https://www.legislature.mi.gov/Laws/MCL?objectName=mcl-169-259), Michigan Legislature, accessed November 16, 2026.
- [MCL 168.932f](https://www.legislature.mi.gov/Laws/MCL?objectName=mcl-168-932f), Michigan Legislature, accessed November 16, 2026.
- [Wis. Stat. 11.1303](https://docs.legis.wisconsin.gov/statutes/statutes/11/xiii/1303), Wisconsin Legislature, accessed November 16, 2026.
- [Fla. Stat. 106.145, Use of artificial intelligence](https://www.leg.state.fl.us/statutes/index.cfm?App_mode=Display_Statute&URL=0100-0199/0106/Sections/0106.145.html), Florida Legislature, accessed November 16, 2026.
- [Utah Code 20A-11-1104, Disclosure of synthetic media](https://le.utah.gov/xcode/Title20A/Chapter11/20A-11-S1104.html), Utah Legislature, accessed November 16, 2026.
- [New York Election Law 14-106 (PDF)](https://legislation.nysenate.gov/pdf/laws/ELN14-106), New York State Senate Open Legislation, accessed November 16, 2026.
- [RCW 29B.35, Synthetic media in campaign advertising](https://app.leg.wa.gov/RCW/default.aspx?cite=29B.35&full=true), Washington State Legislature, accessed November 16, 2026; see also the [chapter 42.62 RCW disposition table](https://app.leg.wa.gov/RCW/dispo.aspx?cite=42.62) and [RCW 29B.10.220](https://app.leg.wa.gov/RCW/default.aspx?cite=29B.10.220).
- [Minnesota Statutes 609.771, Use of deep fake technology to influence an election](https://www.revisor.mn.gov/statutes/cite/609.771), Minnesota Office of the Revisor of Statutes, accessed November 16, 2026.
- [SB 751, relating to fabricating a deceptive video with intent to influence an election (enrolled)](https://capitol.texas.gov/tlodocs/86R/billtext/html/SB00751F.htm), Texas Legislature Online, 2019.
- [A.R.S. 16-1024, Deepfakes; candidates](https://www.azleg.gov/ars/16/01024.htm), Arizona State Legislature, accessed November 16, 2026; and [A.R.S. 16-937](https://www.azleg.gov/ars/16/00937.htm).
- [S.23 (Act 75), bill status](https://legislature.vermont.gov/bill/status/2026/S.23), Vermont General Assembly, accessed November 16, 2026; [Act 75 as enacted (PDF)](https://legislature.vermont.gov/Documents/2026/Docs/ACTS/ACT075/ACT075%20As%20Enacted.pdf).
- [MCA 13-35-802](https://mca.legmt.gov/bills/mca/title_0130/chapter_0350/part_0080/section_0020/0130-0350-0080-0020.html), Montana Legislature, Montana Code Annotated 2025.
- [*Kohls v. Bonta*, No. 2:24-cv-02527, Order and Final Judgment and Permanent Injunction as to AB 2655, ECF No. 98 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.caed.453046/gov.uscourts.caed.453046.98.0.pdf), U.S. District Court for the Eastern District of California, August 20, 2025.
- [*Kohls v. Bonta*, Stipulation and Order Regarding AB 2655 Enforcement, ECF No. 100 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.caed.453046/gov.uscourts.caed.453046.100.0.pdf), U.S. District Court for the Eastern District of California, filed August 29, 2025.
- [*Kohls v. Bonta*, Order Granting Plaintiffs’ Motion for Summary Judgment (AB 2839), ECF No. 101 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.caed.453046/gov.uscourts.caed.453046.101.0.pdf), U.S. District Court for the Eastern District of California, August 29, 2025.
- [*Kohls v. Bonta*, Notice of Appeal, ECF No. 106 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.caed.453046/gov.uscourts.caed.453046.106.0.pdf), U.S. District Court for the Eastern District of California, September 19, 2025.
- [*The Babylon Bee, LLC v. Bonta*, No. 25-6138, docket](https://www.courtlistener.com/docket/71503636/the-babylon-bee-llc-et-al-v-bonta-et-al/), U.S. Court of Appeals for the Ninth Circuit, via CourtListener RECAP.
- [*The Babylon Bee, LLC v. Lopez*, No. 1:25-cv-00234, Order Granting Plaintiffs’ Motion for Summary Judgment, ECF No. 47 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.hid.174048/gov.uscourts.hid.174048.47.0.pdf), U.S. District Court for the District of Hawaii, January 30, 2026.
- [*Kohls v. Ellison*, No. 25-1300 (PDF)](https://ecf.ca8.uscourts.gov/opndir/26/02/251300P.pdf), U.S. Court of Appeals for the Eighth Circuit, February 9, 2026.
- [*Kohls v. Ellison*, No. 25-1300, order denying rehearing (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.ca8.111041/gov.uscourts.ca8.111041.00805486802.0.pdf), U.S. Court of Appeals for the Eighth Circuit, March 31, 2026, via CourtListener RECAP.
- [*Accountability in State Government v. Knudsen*, No. 6:26-cv-00038, Order Granting Plaintiffs’ Motion for Preliminary Injunction, ECF No. 31 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.mtd.84819/gov.uscourts.mtd.84819.31.0.pdf), U.S. District Court for the District of Montana, September 16, 2026; [docket](https://www.courtlistener.com/docket/73306503/accountability-in-state-government-v-knudsen/) via CourtListener RECAP, last entry November 10, 2026.
- [*Poitras v. Clark*, No. 2:26-cv-00234, Opinion and Order, ECF No. 40 (PDF)](https://storage.courtlistener.com/recap/gov.uscourts.vtd.41405/gov.uscourts.vtd.41405.40.0.pdf), U.S. District Court for the District of Vermont, September 4, 2026; [docket](https://www.courtlistener.com/docket/73590007/poitras-v-clark/) via CourtListener RECAP, last entry October 26, 2026.
- [Fraudulent Misrepresentation of Campaign Authority, interpretive rule, Notice 2024-24 (89 FR 78785)](https://www.federalregister.gov/documents/2024/09/26/2024-21983/fraudulent-misrepresentation-of-campaign-authority), Federal Election Commission, Federal Register, September 26, 2024.
- [Artificial Intelligence in Campaign Ads, notification of disposition of petition for rulemaking, Notice 2024-23 (89 FR 78826)](https://www.federalregister.gov/documents/2024/09/26/2024-21979/artificial-intelligence-in-campaign-ads), Federal Election Commission, Federal Register, September 26, 2024.
- [Artificial Intelligence in Campaign Ads, notice of availability (88 FR 55606)](https://www.federalregister.gov/documents/2023/08/16/2023-17547/artificial-intelligence-in-campaign-ads), Federal Election Commission, Federal Register, August 16, 2023.
- [Disclosure and Transparency of Artificial Intelligence-Generated Content in Political Advertisements, proposed rule (89 FR 63381)](https://www.federalregister.gov/documents/2024/08/05/2024-16977/disclosure-and-transparency-of-artificial-intelligence-generated-content-in-political-advertisements), Federal Communications Commission, Federal Register, August 5, 2024.
- [MB Docket No. 24-211, proceeding detail](https://www.fcc.gov/ecfs/search/docket-detail/24-211), FCC Electronic Comment Filing System, accessed November 16, 2026.
- [Declaratory Ruling, Implications of Artificial Intelligence Technologies on Protecting Consumers from Unwanted Robocalls and Robotexts, CG Docket No. 23-362, FCC 24-17 (PDF)](https://docs.fcc.gov/public/attachments/FCC-24-17A1.pdf), Federal Communications Commission, released February 8, 2024.
- [Steve Kramer, Forfeiture Order, FCC 24-104 (PDF)](https://docs.fcc.gov/public/attachments/FCC-24-104A1.pdf), Federal Communications Commission, adopted September 26, 2024, released September 30, 2024.
- [S. 1213, Protect Elections from Deceptive AI Act, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/s/BILLSTATUS-119s1213.xml), U.S. Government Publishing Office; text: [S. 1213 as introduced](https://www.govinfo.gov/content/pkg/BILLS-119s1213is/html/BILLS-119s1213is.htm).
- [H.R. 5272, Protect Elections from Deceptive AI Act, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr5272.xml), U.S. Government Publishing Office.
- [H.R. 9922, AI Transparency in Elections Act of 2026, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr9922.xml), U.S. Government Publishing Office; text: [H.R. 9922 as introduced](https://www.govinfo.gov/content/pkg/BILLS-119hr9922ih/html/BILLS-119hr9922ih.htm).
- [S. 5135, AI Ads Act, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/s/BILLSTATUS-119s5135.xml) and [H.R. 9985, AI Ads Act, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr9985.xml), U.S. Government Publishing Office.
- [S. 4774, FAIR Elections Act of 2026, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/s/BILLSTATUS-119s4774.xml), U.S. Government Publishing Office.
- [H.R. 10814, REAL Political Advertisements Act, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr10814.xml) and [H.R. 10841, Securing Elections From AI Deception Act, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr10841.xml), U.S. Government Publishing Office.
- [S. 4591, NO FAKES Act of 2026, bill status (XML)](https://www.govinfo.gov/bulkdata/BILLSTATUS/119/s/BILLSTATUS-119s4591.xml), U.S. Government Publishing Office.
- [Political content](https://support.google.com/adspolicy/answer/6014595?hl=en), Google Advertising Policies Help, accessed November 16, 2026.
- [Ads about Social Issues, Elections or Politics](https://transparency.meta.com/policies/ad-standards/SIEP-advertising/SIEP/), Meta Transparency Center, accessed November 16, 2026.

## Common questions

**Do political ads have to disclose AI?**

Only where a state law requires it, and only for the ads that law covers. As of November 16, 2026, no federal statute requires an AI label on political ads, the FEC has declined to open an AI rulemaking, and we found no adoption of the FCC's 2024 proposal for broadcast disclosures. States such as California, Michigan, Wisconsin, Florida, Utah, and New York set their own label text and exceptions, and others, such as Arizona and Vermont, bar a deceptive deepfake within 90 days before an election unless it carries the required disclosure.

**Are state election deepfake laws enforceable?**

They apply unless a court has blocked them, and each court order reaches a specific law and specific parties. California's AB 2839 is permanently enjoined as to the named plaintiffs and AB 2655 as to X Corp. and Rumble (the state appealed the AB 2655 judgment), and Hawaii's Act 191 is permanently enjoined. California rewrote the AB 2839 section by an urgency statute, AB 502, in September 2026; we found no court order on the new text as of November 16, 2026. In September 2026, federal courts issued preliminary injunctions protecting the named plaintiffs against Montana's law and, in Vermont, one plaintiff against one investigative demand and any Act 75 enforcement over one video.

**Does the FEC regulate AI in campaign ads?**

Only through the existing ban on fraudulent misrepresentation of campaign authority. On September 26, 2024, the FEC published its decision not to open an AI rulemaking and an interpretive rule, effective that day, that 52 U.S.C. 30124 and 11 CFR 110.16 apply irrespective of the technology used, including AI. The FEC states that the interpretive rule does not bind the Commission or any member of the public, and it is not a disclosure requirement.

