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

A study from Convovo, a conversational AI research studio led by Cornell professor Tapan Parikh, asked GPT-6.1 Sol and Grok 4.7 through their developer APIs to name one candidate in nearly 500 midterm races. Grok picked the Republican in 92 percent of answers. GPT picked the Democrat in 97 percent. They agreed less than 10 percent of the time.

The consumer apps refused to answer the same question, and OpenAI stresses that point: the API test used a prompt explicitly demanding a recommendation, which isn't how most people use ChatGPT. Fair enough. But developers build products on these APIs, Parikh says simple jailbreaks got the consumer apps to answer too, and the midterms are November 3.

We don't think the headline number is the real story. Almost nobody should be asking a chatbot who to vote for, and Pew says only 8 percent of Americans use chatbots for election information at all. The real story is that these systems carry values, and those values show up in quieter ways: which issues they search for, which candidates they look up, and which sources they trust.

The study's breakdown is where it gets useful. GPT searched more for healthcare and abortion positions, and sometimes looked up only the Democrat. Grok searched for immigration, taxes, crime and energy, and frequently pulled from conservative sources, including a Christian conservative voter guide in 39 percent of answers, Turning Point Action's site in 23 percent, and X in 41 percent. GPT leaned on Wikipedia, the Associated Press, Axios and Reddit, several of which have content deals with OpenAI. A user asking a neutral-sounding question about what matters in a race would get a different framing from each, and probably never notice.

The search-off test, run on a small sample of 30 races, is the most revealing part. Without web search, GPT still picked the Democrat in 24 of 30, which suggests the lean is in the model itself. Grok mostly refused to choose, and in the three races where it did, it picked the Democrat. Parikh's read is that Grok's lean comes from how it searches, not what it learned. The researchers also found that every Grok API request carried roughly 1,240 tokens of added input they didn't write and couldn't see. We can't say what those tokens contain, and neither can they. That opacity is the problem.

I want to be careful here. This is one study, using a forcing prompt, with a 30-race follow-up. It doesn't prove intent by either company, and OpenAI says its products are designed to be objective. But both companies sit close to politics: SpaceXAI's CEO has given upwards of $445 million to GOP midterm candidates, and Sam Altman has historically donated to Democrats, though he says he isn't funding 2026 races. When the people at the top are this politically visible, the bar for transparency should be higher, not lower.

Where this leads if it scales is the part that worries us. As more people use AI to research issues instead of reading candidate pages or local news, small differences in emphasis and sourcing add up. That kind of influence is harder to spot than an endorsement, and the Center for AI Safety has already warned that models can covertly push users toward one side. We're pro-AI as a research tool, including for civic life. That only works if the tools disclose what's shaping their answers.

The study tested developer APIs, which companies use to build AI into their own products, rather than the consumer ChatGPT and Grok apps. OpenAI said the API test doesn't reflect the ordinary ChatGPT experience. SpaceXAI did not respond to TechCrunch in time for publication. Earlier Grok models were found to consult Elon Musk's views before answering controversial questions.