Texas Senate Race Shows AI Deepfakes Now Standard Campaign Tool
Ken Paxton and John Cornyn both deployed AI-generated attack ads in their primary battle, highlighting how synthetic media has become routine in political warfare.
Artificial intelligence-generated political attack ads have moved from experimental novelty to standard campaign arsenal in Texas, according to reporting from the Austin American-Statesman.
During the Republican primary for U.S. Senate, Texas Attorney General Ken Paxton released an AI-generated video falsely depicting incumbent Sen. John Cornyn dancing with Democratic Rep. Jasmine Crockett. The deepfake spliced fabricated footage with real clips of Crockett discussing bipartisan work with Cornyn, creating a misleading narrative about their political relationship.
Cornyn responded with his own AI-generated attack, showing Paxton in a 1960s convertible set to a cover of "Love Shack" — a reference to allegations of extramarital affairs that contributed to Paxton's wife filing for divorce. The exchange marked one of the first major Republican primaries where both candidates deployed synthetic media as core campaign tactics.
Why it matters
The normalization of deepfakes in high-profile races signals a fundamental shift in political communication. When both establishment and insurgent candidates treat AI-generated content as routine campaign tools rather than experimental tactics, the technology has crossed a threshold. This creates new challenges for voters trying to distinguish authentic footage from fabricated content, and for platforms and regulators attempting to maintain information integrity during elections.
State Law Offers Limited Protection
Texas enacted legislation in 2019 prohibiting deepfake videos in elections within 30 days of voting. However, the law explicitly excludes federal races like the Senate contest. State Sen. Bryan Hughes, who authored the measure, told the Statesman he designed the restriction to balance free speech protections with voter protection, reasoning that earlier deployment allows time for corrections.
The 30-day window means Paxton's March primary deepfake and his subsequent attack on Democratic nominee James Talarico — which showed the seminarian purportedly editing biblical passages to claim God created six biological sexes — both fall outside the law's scope.
Effectiveness Remains Unclear
Whether deepfakes actually sway voters remains uncertain. Paxton launched his Cornyn-Crockett video before the March primary, where Cornyn won the most votes but failed to avoid a runoff. Cornyn's counterpunch came during the runoff campaign, which Paxton won decisively. The lopsided runoff result suggests factors beyond AI-generated content drove the outcome.
Still, campaigns continue investing in the technology, indicating operatives believe synthetic media offers strategic value even if direct electoral impact is difficult to measure.
The Historical Context
Political satire predates modern technology by centuries. The Statesman notes that 18th-century cartoons depicted Thomas Jefferson as a rooster-like creature burning the Constitution — imagery clearly understood as commentary rather than documentation. The challenge with AI-generated video is that production quality can make fabricated content difficult to distinguish from authentic footage, especially for voters scrolling quickly through social media feeds.
These details were first reported by John C. Moritz in the Austin American-Statesman.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call
