Forced to pick, Grok’s AI model chooses Republicans. OpenAI’s picks Democrats.

A Convovo study found sharply different partisan recommendations from OpenAI’s GPT-6.1 Sol and SpaceXAI’s Grok 4.7 when their developer APIs were forced to name a candidate in nearly 500 U.S. midterm races. Grok chose Republicans in 92% of responses, while GPT chose Democrats in 97%, and the models agreed in fewer than 10% of races. The study comes less than two weeks before the November 3 midterms. Consumer versions of ChatGPT and Grok declined to tell users whom to vote for, although Cornell professor Tapan Parikh said further tests used simple jailbreak prompts to obtain recommendations. The models used different search patterns and sources. Grok frequently cited conservative sources, including a Christian conservative voter guide in 39% of answers, Turning Point Action in 23%, and X in 41%, while also citing Ballotpedia, OpenSecrets, and VoteSmart. GPT relied heavily on Wikipedia, the Associated Press, Axios, and Reddit. GPT more often searched candidates’ healthcare and abortion positions; Grok focused more on immigration, taxes, crime, and energy. With internet search disabled in a 30-race sample, GPT selected Democrats in 24 races, one Republican, and declined five. Grok declined in 27 races and selected Democrats in the three remaining cases. Parikh said those results suggest GPT’s lean is embedded in its model weights, while Grok’s results are driven more by its search process and sources. The study also found that Grok’s API requests included roughly 1,240 additional tokens that researchers could not view or remove.