Republican candidates and affiliated groups account for 88% of national spending on identified AI political advertisements, while federal records document more than $92,000 in expenditures opposing Michigan Democratic Senate candidate Abdul El-Sayed.
LANSING, Mich. — Artificial intelligence-generated political advertisements are becoming a significant factor in Michigan’s U.S. Senate race, with a Republican-aligned political action committee reporting more than $92,000 in expenditures opposing Democratic candidate Abdul El-Sayed through an advertisement featuring fabricated scenes involving the candidate, family members and a political commentator.
The Michigan advertisement is part of a national surge in AI-generated political advertising ahead of the November 3 midterm elections.
According to research released October 15 by the Wesleyan Media Project, political candidates and organizations have spent more than $137 million on 220 identified advertisements containing AI-generated or AI-enhanced material.
Republican candidates and affiliated organizations sponsored 81% of those advertisements and accounted for 88% of spending. Democratic candidates and affiliated organizations sponsored approximately 12% of the advertisements.
Researchers also found that only 27% of the identified advertisements contained disclosures informing viewers that artificial intelligence had been used.
The findings raise questions about whether voters can distinguish authentic political footage from fabricated images, voices and video — and whether Michigan’s political advertising disclosure requirements provide adequate protection.
Federal Records Identify $92,187 In Spending Against Michigan Senate Candidate
The Front Line Action, a Republican-aligned super PAC, produced an AI-generated advertisement titled Family Values targeting El-Sayed.
According to an October 19 Axios investigation, the advertisement depicts fictional scenes involving El-Sayed, members of his family and political commentator Hasan Piker.
One scene features a fabricated family photograph. Another shows an AI-generated board game containing inflammatory references to the September 11 terrorist attacks.
The scenes were artificially generated and do not document actual events involving the individuals depicted.
A September 14 Federal Election Commission filing identifies two expenditures reported by Front Line Action opposing El-Sayed:
- $80,000 to Armada Strategies LLC for media.
- $12,187.50 to Madison McQueen LLC for media production.
The combined expenditures totaled $92,187.50.
The filing lists September 11 as the expenditure or obligation date and September 12 as the public distribution date.
These figures represent reported expenditures opposing El-Sayed, not a comprehensive accounting of all AI advertising expenditures in Michigan.
Importantly, the advertisement contains both written and spoken disclosures that artificial intelligence was used.
Whether those disclosures satisfy every applicable requirement of Michigan law has not been independently established.
Who Is Financing The AI Attack Advertisements?
The Front Line Action registered with the Federal Election Commission on June 12, 2026, as an independent-expenditure-only political committee, commonly called a super PAC.
These organizations can raise and spend unlimited amounts supporting or opposing candidates, provided their expenditures are not coordinated with the candidates’ campaigns.
According to the Federal Election Commission’s official financial records, Front Line Action reported the following activity through September 30:
- Total receipts: $984,500.
- Total disbursements: $898,153.
- Individual contributions: $427,000.
- Contributions from other political committees: $500,000.
The committee’s summary records do not establish the ultimate source of every contribution passing through another political organization.
Tracing those funds would require examining additional itemized contribution records.
Other Michigan Campaigns Have Used AI Political Content
AI-generated political content has also appeared in Michigan’s gubernatorial campaign and in advertising distributed by Republican Senate candidate Mike Rogers.
An August MLive investigation documented an AI-generated video posted by Rogers targeting El-Sayed.
The video depicted a Trojan horse crossing the Mackinac Bridge carrying a sign reading “definitely not communism.”
MLive reported that the video did not appear to contain an AI disclosure and that Rogers’ campaign did not respond to a question about the missing disclaimer.
However, whether the video qualified as a regulated paid advertisement under Michigan law requires additional examination.
Michigan’s Republican gubernatorial primary also featured AI-generated campaign videos.
In June, Republican candidate Aric Nesbitt posted a fictional slasher-film-style video featuring Michigan Gov. Gretchen Whitmer, California Gov. Gavin Newsom, Minnesota Gov. Tim Walz and Michigan Secretary of State Jocelyn Benson.
Republican gubernatorial candidate Perry Johnson also published an AI-generated video depicting himself pursuing rival John James over his refusal to debate.
These examples demonstrate how campaigns are using artificial intelligence for different purposes, including satire, animation and political criticism.
Not every use of AI is deceptive, and not every AI-generated campaign video is subject to identical legal requirements.
Michigan Law Requires Disclosures On Certain AI Political Advertisements
Michigan lawmakers enacted legislation in 2023 requiring disclosures for certain political advertisements generated wholly or substantially using artificial intelligence.
Public Act 263 of 2023, which took effect February 13, 2024, established requirements for qualifying paid political advertisements containing AI-generated images, audio or video.
Under Michigan Compiled Laws Section 169.259, covered video advertisements containing audio generally must include:
- A visual disclosure appearing for at least four seconds.
- An audible disclosure lasting at least three seconds.
- Disclosure language meeting statutory readability and audibility requirements.
The law includes exceptions, including one for qualifying satire or parody.
Violations are civil infractions carrying maximum fines of $250 for a first violation and $1,000 for a second or subsequent violation.
Each noncompliant qualifying advertisement distributed or aired can constitute a separate violation.
The relatively modest penalties raise questions about whether they provide sufficient deterrence in political campaigns spending millions of dollars on advertising.
However, determining whether a specific advertisement violates Michigan law requires examining its content, distribution, disclosures and applicable exemptions.
Nearly Three-Quarters Of Identified AI Political Ads Lack Disclosures
The Wesleyan Media Project found that only 27% of the 220 identified AI-generated or AI-enhanced political advertisements contained disclosures.
Approximately 73% did not contain disclaimers identified by researchers.
That does not mean every advertisement without a disclosure violated election laws. State requirements vary, and some AI-generated content falls outside disclosure requirements.
Researchers identified AI-generated political advertisements in 34 states.
They acknowledged that their estimates likely understate actual AI advertising activity because some artificially generated material is difficult to detect.
For comparison, Wesleyan researchers identified less than $50,000 in spending on AI-generated political advertisements during the 2024 election cycle.
The $137 million represents estimated spending on advertisements incorporating AI, not the amount paid to AI technology companies to produce them.
Michigan Senate Advertising Spending Reaches $181 Million
The rise of AI-generated political content comes during an expensive Michigan election cycle.
According to an October 15 Axios report, Michigan’s U.S. Senate contest had attracted approximately $181 million in political advertising spending.
Republican candidates and affiliated organizations accounted for approximately 58% of that spending.
The $181 million includes conventional political advertising and should not be confused with expenditures specifically involving AI-generated advertisements.
Michigan-specific AI advertising spending has not been comprehensively established.
The Front Line Action filing provides one documented example of spending associated with an AI-generated political attack targeting a Michigan candidate.
Has Michigan Enforced Its AI Political Advertising Law?
An important question remains unanswered: How actively is Michigan enforcing its AI political advertising disclosure requirements?
MITechNews has not independently established whether Michigan authorities have brought enforcement actions under the disclosure provisions during the 2026 election cycle.
Nor has the publication established whether complaints involving the advertisements targeting El-Sayed have been filed with state authorities.
That does not mean no complaints or investigations exist.
Michigan Attorney General Dana Nessel’s office and Secretary of State Jocelyn Benson’s office could provide additional information about complaints, investigations and enforcement.
Other questions include how the law applies to out-of-state political committees and whether existing penalties provide an effective deterrent.
What Michigan Voters Should Know Before November 3
With Election Day approaching, artificial intelligence presents voters with a new challenge.
Photographs, videos and audio recordings that appear authentic may have been created or substantially modified using AI.
Not every AI-generated political advertisement is misleading. The technology can also be used for animation, satire and conventional production effects.
But realistic depictions of political candidates saying or doing things that never happened can make it harder for voters to distinguish authentic information from fabricated content.
Michigan’s disclosure law provides one mechanism for informing voters when certain political advertisements use AI-generated material.
Whether those requirements are consistently followed — and how effectively they are enforced — remains an important question.
For Michigan voters, an apparently realistic political video is no longer, by itself, proof that the events it depicts actually occurred.
MITechNews will continue examining AI-generated political advertising in Michigan, including campaign spending, disclosure requirements and enforcement ahead of the November 3 election.
