Surprise! These top AI bots are politically left wing

A systems engineer prompted a slew of AI bots to take the Political Compass test online to determine where their answers landed on the political spectrum between the right and the left. To absolutely no one’s surprise, most of the bots fell onto the left side of the map with only one in the center, leaning right. Any guesses which one?

The Political Compass is an exam comprising 62 propositions designed to reveal the political ideology of the person taking the test. Instead of people, though, a systems engineer based in Denmark who goes by the name Zapador had a whole host of AI bots take the test. Among the frontier candidates were ChatGPT by OpenAI, Claude by Anthropic, Gemini by Google, Grok by SpaceXAI, plus many more.

AI platforms with political opinions will have an effect on the conclusions people draw and the opinions they form.

To ensure that the results were as accurate as possible, Zapador had each bot take the test “at least five times” before plotting their average scores on a graph. The results are included below.

AI Political Compass Test Results/aipolcom.net

The numbers don’t lie

As you can see, most of the AI bots fell well within the left libertarian category. In fact, only one AI bot differed significantly. The two gray dots on the center libertarian line and the single dot in the right libertarian section represent three versions of Grok — Grok 4.3 is most right-leaning, Grok 4.5 is next, and Grok 4.6 is closest to the line.

Looking at the other side, Google Gemini 3.5 Flash-Lite with no reasoning was the farthest left out of the mainstream options tested, followed by open-weight models Gemma 4 31B and GPT-OSS 120B.

Needless to say, if you’re looking for an AI bot that takes a balanced-to-right-leaning approach to its responses, Grok’s latest models are the only options on the table.

Unfortunately, when looking at global market share based on website traffic, Grok only accounts for 2.4% of the AI market. ChatGPT sits firmly in the lead at 53.9%, followed by Google Gemini at 27.9% and Anthropic Claude at 9.2%. That means most people use left-leaning AI bots in their daily lives, whether they trust it to gather information, complete tasks, or offer input on projects at work.

RELATED: The next AI race isn’t about smarter machines. It’s about human experience.

R.Satish Babu/Getty Images

Biased by default

It’s no mystery how things ended this way. Most of the Big AI companies that control today’s leading AI platforms are leftist organizations with leftist employees. For example, Sam Altman made donations to Kamala Harris’ campaign leading up to her presidential run against Donald Trump, while Dario Amodei reportedly made donations to President Joe Biden, among other Democrats.

It’s also not a coincidence that the one right-leaning chatbot, Grok, is owned by a center-to-right-leaning innovator, Elon Musk.

The big question is whether or not AI chatbots’ political bias is intentional or a side effect of the source material — something we may never know for sure. Still, the implications of a left-aligned AI army that spurns right-leaning values could have significant implications on society, especially as AI sinks deeper into the fabric of the modern world.

Consequences of biased AI

A barrage of left-aligned AI platforms could impact people in dramatic, perhaps unexpected ways.

“There is no question that in a society where people increasingly rely on AI for education and information — to tell them who to vote for and what opinions to have on a variety of topics — that AI platforms with political opinions will have an effect on the conclusions people draw and the opinions they form,” Perry Metzger, chairman of Alliance for the Future, told Blaze Media.

This is true in several facets of life — as students increasingly use AI to mold their education, businesses rely on AI to influence their companies, and voters turn to AI for vital information before an election.

Going further, Metzger expressed his concerns around Chinese-made open-source AI models, like DeepSeek and Kimi, moving in as cheap alternatives to American frontier models, both here in the States and in other nations around the world. Unlike American models that still offer information openly, he noted, Chinese models have their own opinions around topics that question the Chinese government, rightful territory, and more, skewing outputs in a way that could further “twist society.”

As for what to do about it? More government isn’t the answer. Any well-intentioned guidance to improve the training data or minimize bias by the Trump administration could be weaponized by the next Democrat president. Stronger governance over AI outputs would also face constitutional hurdles.

Ultimately, the best thing we can do right now, as a society, is to be aware that these models are biased toward the left and to engage with them with this information top of mind.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!