Seven weeks before the 2026 midterm elections, voters in Texas can no longer trust what they see on their screens, and the rest of America is not far behind.
AI-generated campaign ads have moved from novelty to standard practice in the Lone Star State. According to The Texan, Republican candidates in particular have embraced synthetic media as a tool of political communication, with some videos carrying disclaimers only in small text confirming that AI created them.
The marquee example is the Texas Senate race. As reported by the Austin American-Statesman, Attorney General Ken Paxton targeted Senator John Cornyn in the Republican primary with an AI-generated ad falsely showing the incumbent dancing with a Democrat. Deepfakes have continued into the general election, with Axios reporting that Paxton released a realistic AI ad targeting the Democratic nominee.
How Bad Is the AI Deepfake Problem in Texas?
Extremely bad, by every available account. AI-generated content has saturated the Texas Senate race, with deepfakes depicting candidates in fabricated scenarios ranging from luxury vacations to staged attacks on religious faith. Watchdog groups and news outlets describe it as the most synthetic media saturated contest of the 2026 cycle.
The 19th reported that AI content targeting Democratic nominee James Talarico has shown him in a dress singing about gender affirming care, dressed as a surgeon threatening a child on an operating table, and defacing a Bible. Talarico is a religious scholar, which makes the fabricated Bible imagery a direct strike at his core identity.
On the Republican side, The Texan described a deepfake of Paxton showing the attorney general on luxurious vacations, employees hiding under desks while reporting him to the FBI, and shredding documents. Former House Speaker Dade Phelan warned in an August 15 post that as AI evolves, it becomes increasingly difficult for voters to differentiate between what is real and what is manufactured.
Lieutenant Governor Dan Patrick is the latest official caught in the crossfire. KSS Radio reported on September 14 that a newly released campaign ad from Patrick ignited a multi-front firestorm over its AI content.
What Is Congress Doing About AI Political Ads?
Congress is moving, but slowly. The AI Transparency in Elections Act of 2026, introduced by Democratic Representative Joe Morelle on July 23, would require political advertisements to include clear disclosures when images, audio, or video have been substantially produced using generative AI. Separate bills targeting candidate impersonation and fraudulent fundraising are also advancing.
The problem, as Security Boulevard argues, is that regulation has not worked so far. A cloned candidate voice, spoofed campaign account, fake PAC, or lookalike donation domain can reach voters before traditional legal and communications processes have time to react. Election security has permanently evolved beyond protecting the machines that tally votes.
The publication makes a point every campaign manager should internalize: a political campaign operates like a Fortune 500 enterprise, with a brand identity, highly visible executives, a sprawling vendor supply chain, and a massive base of digital consumers. Threat actors hijack that brand to execute devastating social engineering attacks. Campaigns that fail to monitor their digital footprint are leaving their candidates exposed, which is why disciplined voter contact operations like HyperPhonebank matter more than ever. Direct conversation with a real voter cuts through synthetic noise in a way no algorithm can fake.
Can Voters Still Tell What Is Real?
Increasingly, no, and the consequences extend beyond any single race. Axios reports that Americans' trust in politicians and institutions has been tanking for years, and political tech experts warn that AI campaign ads could deepen that erosion. When every video might be fake, authentic messages lose power too.
The information environment itself is shifting. The New York Times reported that voters who turned to AI chatbots a day before Alaska's primary election last month may have received information that was incomplete or tuned to who they were. MIT researchers Chara Podimata, Adam Berinsky, and Charles Stewart III are now studying how AI answers differ across demographics, political identities, and geographies.
Microsoft announced on September 15 new efforts to help voters navigate election information in the age of AI, an acknowledgment that the platforms distributing political content now shoulder part of the burden of defending it.
The Threat Goes Far Beyond Campaign Ads
Anthropic's September 2026 threat intelligence report documents influence operations using AI-generated personas at scale. In one Kenya focused network, actors used AI profile photos, invented biographies for fake reporters, and fabricated political spokespeople to push hashtags like #PowerReliefKE and to spread claims that the United Opposition coalition was breaking apart ahead of the 2027 general election. The same report describes impersonation of real institutions, including a state spokesperson and a human rights organization, along with forged government documents.
Election infrastructure is not immune either. The Christian Science Monitor reported that a vulnerability in Georgia, known for years, is drawing fresh alarm because AI can now do the technical work of exploiting it. University of Michigan voting technology specialist J. Alex Halderman called it a wake-up call, noting the flaw could let someone see how others voted and potentially match voters to their ballots, threatening the constitutional right to ballot secrecy.
Meanwhile, the politics of AI itself is becoming a campaign issue. NBC News reports that backlash against the nationwide data center buildout is threatening to shape the 2026 midterms, and the bipartisan strategist team of Josh Vlasto and Zac Moffatt, atop the pro-AI super PAC Leading the Future, says these fights are merely in the early innings.
What Should Campaigns Do Right Now?
First, build trust capital before an attack hits. As The Economy notes, trust capital is becoming the defining asset of the AI era, because campaigns with established credibility survive synthetic smears better than those starting from zero.
Second, verify everything before amplifying it, and demand the same discipline from surrogates. Campaigns should also monitor the information environment the way they monitor polling. Tools like HyperPulse, which applies AI to synthetic polling and prediction, give strategists a way to understand how voters are actually processing a media landscape saturated with fabricated content.
Third, invest in communication channels voters still trust. Live phone conversations, door knocks, and community presence cannot be deepfaked at scale. Campaigns can explore the full range of services built for exactly this environment, and track the latest developments through our phone banking coverage.
The 2026 midterms will be remembered as the cycle when synthetic media stopped being a sideshow. The campaigns that win in November will be the ones that treated authenticity as infrastructure, not as an afterthought.