Alleged Leftward Bias in Leading AI Chatbots

The conservative watchdog Media Research Center (MRC) released studies claiming the three most used AI chatbots—OpenAI’s ChatGPT, Google’s Gemini, and Anthropic’s Claude—show a consistent left-wing slant. Cited in Wynton Hall’s bestseller *CODE RED*, the analysis says the systems suppress conservative viewpoints while amplifying liberal ones. MRC notes Google’s multi-year federal contracts and that President Donald Trump’s 2025 AI Action Plan requires AI used in procurement to be free of ideological bias.

Breitbart

MRC’s tests highlight several examples. In January, ChatGPT answered that “nobody named Charlie Kirk was ever assassinated” and found “no credible evidence.” Claude reportedly refused to embed the U.S. First Amendment into its policies. When asked about Maine Democratic Senate candidate Graham Platner, Gemini focused on his Time-magazine cover, polling leads and endorsements, mentioning scrutiny only briefly and omitting the Fox News Digital report on his controversial Reddit posts. ChatGPT framed the issue as Republican attack ads, while Claude cited Platner’s past joke about becoming a communist without noting newer allegations.

BreitbartMedia Research Center Exposes Leftist Bias in AI Chatbots – NewsFire.GRBreitbart

President David Bozell warned, “We’re watching the next phase of media bias unfold in real time,” and added, “Silicon Valley’s shiny new toys can no longer be considered neutral and cannot be trusted.” Dan Schneider, vice president of MRC’s Free Speech America division, said, “Research from international scientists confirms… these chatbots disproportionately rely on left-wing media and politically charged sources.” The group cites Trump’s AI Action Plan as the standard for unbiased federal AI contracts.

Breitbart

MRC’s conclusions rely on internal tests and proprietary reports, such as Gemini’s 3,400-word “Analytical Assessment of Congressional Rhetoric,” which is not publicly released. No independent audits or peer-reviewed studies are cited, leaving the bias unverified. Methodology details for the ChatGPT “Charlie Kirk” response and Claude’s First-Amendment stance are also missing, creating uncertainty about sample size and query phrasing.

Media Research Center Exposes Leftist Bias in AI Chatbots – NewsFire.GR

• “We’re watching the next phase of media bias unfold in real time,” — David Bozell, President, Media Research Center

Breitbart

• “Silicon Valley’s shiny new toys can no longer be considered neutral and cannot be trusted.” — David Bozell, President, Media Research Center

Breitbart

• “Research from international scientists confirms what we’ve seen firsthand: these chatbots disproportionately rely on left-wing media and politically charged sources to generate their answers,” — Dan Schneider, Vice President, Free Speech America, Media Research Center

Breitbart

• “AI systems must be free from ideological bias” — President Donald Trump, 2025 AI Action Plan

Breitbart

MRC urges tighter oversight of AI procurement and calls on federal agencies to enforce the bias-free clause of Trump’s AI Action Plan. It will keep testing chatbot outputs and publishing examples to shape policy discussions before the next round of government AI contracts.

Breitbart