A new study has uncovered significant political biases in AI models, with SpaceXAI's Grok 4.7 recommending Republican candidates in 92% of races and OpenAI's GPT-6.1 favoring Democrats in 97% of cases. The models agreed on candidates less than 10% of the time across nearly 500 races.
TL;DR
- Grok 4.7 and GPT-6.1 show stark political biases, with Grok favoring Republicans and GPT Democrats.
- The study highlights the importance of understanding AI model biases, especially as more people turn to chatbots for election information.
- The findings could fuel the debate over political bias in AI, with implications for developers and users alike.
What happened
A study by Convovo, led by Cornell University professor Tapan Parikh, found that when forced to choose in upcoming midterm elections, Grok 4.7 recommended the Republican candidate in 92% of its answers, while GPT-6.1 picked the Democrat in 97%. The models agreed on candidates less than 10% of the time across nearly 500 races.
The study asked the models, via their APIs, to name one candidate without hedging or listing both sides. Consumer apps refused to answer, with OpenAI noting that this reflects the models' design to be objective by default. SpaceXAI did not respond to requests for comment.
Parikh noted that the models' biases can show up in less direct questions, such as what topics are most important to consider in an upcoming election. The study also found that the models' sources varied significantly, with Grok favoring conservative sources and GPT leaning on mainstream outlets.
Why it matters
The study highlights the importance of understanding AI model biases, especially as more people turn to chatbots for election information. Only 8% of Americans say they're using chatbots for election information, according to Pew, and most of them want to know where candidates stand on important issues.
For developers, the study underscores the need to be aware of potential biases in the models they use. For investors, it highlights the growing debate over political bias in AI and the potential implications for companies in the space.
The findings also raise questions about the transparency of AI models and the sources they use to make decisions. As the debate over political bias in AI continues, these questions will likely become increasingly important.
Key facts
- Grok 4.7 recommended the Republican candidate in 92% of races.
- GPT-6.1 picked the Democrat in 97% of races.
- The models agreed on candidates less than 10% of the time across nearly 500 races.
- Only 8% of Americans say they're using chatbots for election information, according to Pew.
- Grok frequently surfaced conservative sources, including a Christian conservative voter guide in 39% of answers and X in 41%.
- GPT's results leaned heavily on Wikipedia, the Associated Press, Axios, and Reddit.
- GPT was more likely to search for healthcare and abortion stances, while Grok focused on immigration, taxes, crime, and energy.
- Every request to Grok through xAI's API carried about 1,240 tokens of additional input that the researchers couldn't see or remove.
Context
The study comes less than two weeks before the November 3 midterms, and its findings could fuel the debate over political bias in AI. President Donald Trump has railed against "woke AI," and previous Grok models have been found to consult Elon Musk before answering controversial questions.
OpenAI CEO Sam Altman has historically donated to Democratic candidates, but in recent years has begun supporting more Republicans. This year, he said, he isn't putting money into the 2026 elections.
As the world becomes increasingly charged by identity politics, understanding the political biases of AI models will become increasingly important. For those who want to review each candidate's stance and make their own choices, resources like Vote.gov, Ballotpedia, and VoteSmart can be helpful.