As artificial intelligence platforms continue their race to be the first to reach artificial general intelligence, these same AI platforms are contracting almost exclusively with media outlets with a leftist bias. Left unchecked, the technology can and will shift public perception to the left without most users even realizing it.
MRC President David Bozell warned the Trump administration of this very phenomenon in letters to Office of Management and Budget (OMB) Director Russell Vought and Federal Trade Commission (FTC) Chairman Andrew Ferguson on Wednesday. In the letters, Bozell called attention to top AI companies’ training data agreements with predominantly left-leaning legacy media outlets (59% left-leaning vs. 5% right-leaning). He also asked Vought to clarify whether these contracts violate a recent executive order by President Donald Trump and requested that Ferguson investigate whether the AI firms’ contracts constitute anti-competitive and unfair practices that could distort the marketplace.
OpenAI has opted to train ChatGPT with outlets like The Washington Post, The Atlantic, Financial Times, Vox Media and others. Amazon has similarly made an agreement to use The New York Times and Google has contracted with The Associated Press. News Corp, which has a contract with ChatGPT, is the only right-leaning media group that has this type of contract with a major AI company.
Up until now, these contracts have gone on without a hitch despite President Trump’s recent executive order promoting ideological neutrality. But Amazon, Google, Meta, Microsoft and OpenAI’s 97 agreements with media entities are anything but neutral.
In the letter to OMB Director Vought, Bozell explained: “The Media Research Center’s analysis found that AI platforms are contracting with left-leaning outlets over ten times more often than with outlets rated ‘lean right’ or ‘right’ by AllSides. These contracts—such as those between OpenAI and The Atlantic or The Associated Press—limit competition, reinforce a single ideological perspective, and undermine the diversity of information essential to neutral AI outputs.” [Emphasis added.]
Indeed, OpenAI has more contracts with media outlets than any other top AI company, and an MRC analysis found that only two of its agreements are with media outlets AllSides rates as “right” (Fox News) or “lean right” (the New York Post). Meanwhile, OpenAI has made deals with at least 21 media outlets AllSides deems left-leaning. This raises many legal questions.
In his letter to FTC Chairman Ferguson, Bozell noted that, “Courts have upheld the right to use publicly available material for AI training, suggesting these exclusive agreements are unnecessary and may exist primarily to funnel payments to favored outlets.” [Emphasis added.]
These contracts further bring up antitrust concerns because they “consolidate influence in a few legacy voices [and] reduce market diversity.” Bozell continued, adding, “Big Tech’s selective contracting, combined with its political influence and push for stricter copyright enforcement, appears designed to block new competitors and entrench control over information markets.”
As such, the MRC asked that the FTC to “immediately investigate whether Big Tech’s artificial intelligence firms are engaging in anti-competitive and unfair practices by contracting almost exclusively with ideologically aligned media outlets. ”
Given these AI platforms’ overt reliance on left-leaning sources, MRC also asked that OMB offer guidance clarifying whether the AI companies’ “intentional use of a narrow set of biased sources” in their training data “violates the ideological neutrality requirement in President Donald Trump’s recent Executive Order on AI.” Bozell additionally stressed the need for action noting the “hundreds of millions of dollars recently awarded in federal defense contracts to AI companies.”
As Bozell repeatedly emphasized in his letters, the impact of these contracts is significant given that excluding diversity in artificial intelligence training data would have the effect of biasing its results. It seems both tech CEOs and politicians are well aware of this.
During a roundtable event, former Vice President Kamala Harris said the quiet part out loud. “[T]he machine is taught—and part of the issue here is what information is going into the machine that will then determine—and we can predict then, if we think about what information is going in, what then will be produced in terms of decisions and opinions that may be made through that process.”
Open AI CEO Sam Altman made a similar admission when he testified before the Senate in 2023, saying that “one of my areas of greatest concern” is “[t]he more general ability of [AI] models to manipulate, to persuade, [and] to provide sort of one-on-one interactive disinformation.” Google CEO Sundar Pichai also stated in a 2021 interview with Yahoo! Finance that bias review is “extraordinarily important” and that “studying for bias needs to be done up front” to ensure fairness.
Just last year, Elon Musk similarly acknowledged a growing problem OpenAI’s ChatGPT and Google’s Gemini were experiencing. “The concern I have—and this may seem like a small concern, but I think it’s actually a very big, big issue—is that they are not maximally truth-seeking. They're pandering to political correctness.”
MRC contributor Zackory Langin contributed to this report
Methodology: For the analysis, MRC researchers cross-referenced Amazon, Google, Meta, Microsoft and OpenAI’s 97* agreements with media entities with AllSides’ political bias ratings. AllSides had rated 39 of the media entities. Researchers then tallied how AllSides rated the 39* entities. AllSides rated 21 as “left” or “lean-left,” 14 as “center” and two as “right” or “lean right.”
*OpenAI and Google both have contracts with The Associated Press (which is rated by AllSides) and OpenAI and Microsoft both have contracts with Harper Collins (which is not rated by AllSides). The above numbers reflect that these were counted separately.