Political campaigns across the nation are turning to artificial intelligence to create convincing fake videos of their opponents as the 2026 midterm elections approach, sparking widespread concern about voter manipulation.
A recent example involves Texas State Representative James Talarico, a Democratic U.S. Senate candidate, who appears in a video standing before a Texas flag making controversial statements about domestic terrorism. However, Talarico never recorded this footage – it was entirely computer-generated by the National Republican Senatorial Committee using his old social media posts.
“Radicalized white men are the greatest domestic terrorist threat in our country,” the AI version of Talarico appears to say, followed by “So true. So true” as a voice whispers “white men.” A barely visible disclaimer reading “AI generated” appears in small text at the bottom corner.
This sophisticated fake video represents a growing trend of “deepfake” political advertisements that campaigns are deploying with minimal oversight. Currently, no federal laws restrict AI use in political messaging, leaving only scattered state regulations that remain largely untested.
Social media platforms like Meta and X have moved away from professional fact-checking systems, instead relying on user-generated notes to identify AI content. This shift comes as the technology becomes more accessible and convincing.
The November elections will determine congressional control during the final two years of Republican President Donald Trump’s administration. Democrats appear positioned to potentially regain the House majority while facing steeper challenges in Senate races.
Research published in the Journal of Creative Communications during 2025 revealed that viewers frequently cannot distinguish deepfake videos from authentic content, and these fabricated materials significantly influence public opinion.
Republicans have embraced this technology more extensively than Democrats this election cycle, following the Trump White House’s lead in producing numerous AI-generated videos and internet memes that target various political issues, including protests and foreign policy matters.
NRSC Communications Director Joanna Rodriguez justified their Talarico advertisement, stating Democrats were “panicking after seeing and hearing James Talarico’s own words.”
JT Ennis, representing Talarico’s campaign, responded that while opponents “spend their time making deepfake videos to mislead Texans, we are uniting the people of Texas to win in November.”
Among Democrats, California Governor Gavin Newsom has frequently used AI-generated content to mock Trump, positioning himself as a potential 2028 presidential contender. However, Democratic national campaign organizations have not matched Republican efforts in deploying this technology for midterm races.
Georgia Representative Mike Collins, a Republican challenging Democratic Senator Jon Ossoff, produced a deepfake showing Ossoff apparently stating: “I just voted to keep the government shut down. They say it would hurt farmers, but I wouldn’t know. I’ve only seen a farm on Instagram.”
Collins’ campaign spokesperson defended their approach, saying they “will be at the forefront embracing new tactics and strategies that pierce through lopsided legacy media coverage and deliver our message directly to voters.”
When questioned by the Atlanta Journal-Constitution, Ossoff’s campaign agreed to avoid using deepfakes that misrepresent opponents’ words or actions to deceive voters.
Purdue University professor Daniel Schiff, who has analyzed thousands of deepfake videos, warned about the technology’s potential to undermine democratic institutions and voter confidence.
“I think that the types of damage that we can do to the rigor and credibility of elections and democratic systems – and the ability to misinform people about candidates or social issues – very much risks being supercharged,” Schiff explained.
Despite concerns, political strategists acknowledge AI-generated videos offer persuasive, cost-effective campaign tools when used responsibly. The technology enables political satire in formats ideal for social media sharing.
Twenty-eight states have enacted legislation addressing AI in political advertisements, primarily focusing on disclosure requirements rather than complete prohibitions, according to Public Citizen’s Ilana Beller, who tracks AI-related state legislation.
However, these laws have significant limitations. Many only apply to official campaigns, not individual social media users spreading AI-manipulated content. Research also indicates that disclaimer notices fail to prevent voter persuasion by false advertisements.
The technology’s affordability has enabled smaller campaigns and local political groups to utilize AI tools. In February, the Republican Committee for Loudoun County, Virginia, released three AI-generated attacks against Democratic Governor Abigail Spanberger, who assumed office in January.
One video combined authentic footage of Spanberger’s State of the Union response with AI-generated clips showing her apparently advocating for “commie socialist Marxism, free stuff for illegals, gun grabs and erasing gender norms.”
Neither Spanberger’s office nor the Loudoun County Republican Committee responded to requests for comment.
Some AI advertisements are deliberately absurd. Texas Attorney General Ken Paxton’s primary campaign against Senator John Cornyn featured an obviously fake video of Cornyn dancing with Democratic Representative Jasmine Crockett, accompanied by narration suggesting private cooperation between supposed opponents.
Cornyn’s campaign retaliated with their own AI creation showing Paxton driving with women labeled “Mistress #1” and “Mistress #2,” referencing infidelity allegations against the attorney general.
Neither Paxton nor Cornyn’s campaigns provided responses to comment requests.
This rapid exchange demonstrates how quickly AI-generated attacks are becoming standard campaign practice, despite ongoing concerns about electoral system integrity.
“It’s harmful for politicians and campaigns to continue normalizing this,” Schiff concluded.
