Washington Post Investigation Finds AI Chatbots Have Huge Liberal Bias

LEFT: Elon Musk ((AP Photo/Czarek Sokolowski, File) RIGHT: Sam Altman ( (AP Photo/Markus Schreiber, File)

The Washington Post tested a number of high-profile AI chatbots and found that all of them — including Elon Musk’s Grok — leaned to the left.

“The Washington Post tested the AI models behind OpenAI’s ChatGPT, Google’s Gemini and others using political questions designed by researchers to gauge how chatbots respond to hot-button political issues. The results suggest that chatbots have clear political leanings that can conflict with promises made by the companies behind them,” wrote the Post‘s Kevin Schaul after noting that President Donald Trump has signed an executive order mandating that the bots be “neutral, nonpartisan tools.”

“The model that powers ChatGPT answered nearly every question exclusively with left-leaning arguments and presented only right-leaning positions just once. Google’s Gemini mostly took a both-sides approach, offering both left and right positions in more than 90 percent of its answers,” continued Schaul. “And even AI models marketed as having conservative views, including Elon Musk’s Grok, offered by his company SpaceX, cited left-leaning arguments more often, on average. (The Post has a content partnership with OpenAI.)… AI models were asked answer each of the questions in 30 words, without personalization settings turned on. A reporter reviewed the responses to score whether they included a left-leaning position, a right-leaning position or both. Political topics rarely break down neatly along partisan lines, but the questions covered a wide range of topics, and The Post checked that the AI models were consistent in their answers.”

According to the Post, OpenAI provided only a left-leaning argument to 80% of queries, only a right-leaning argument to 3% of queries, and a more balanced answer to 17%. DeepSeek came in at 70%, 7%, and 23%, Gab put up 50%, 3%, and 47%, and Anthropic’s Claude provided a left-leaning answer in 43% of cases and both sides the remaining 57% of the time.

Grok and Google’s Gemini were by far the most balanced. The former came in at 40%, 33%, and 27%, while Gemini offered both sides 93% of the time and a left-leaning reply the remaining 7%.

A spokeswoman for Google told the Post that “Gemini is designed to provide balanced responses that don’t favor any political ideology,” and a spokesman for Anthropic said that “We train Claude to treat different political viewpoints equally and test extensively for bias before every model launch.”

OpenAI, SpaceX, DeepSeek, and Gab did not respond to the paper’s requests for comment.

New: The Mediaite One-Sheet "Newsletter of Newsletters"
Your daily summary and analysis of what the many, many media newsletters are saying and reporting. Subscribe now!