Missouri Attorney General Andrew Bailey is demanding internal records from major tech companies—Google, Meta, Microsoft, and OpenAI—over allegations that their AI chatbots are distorting facts and exhibiting political bias against President Donald Trump. Bailey warns the companies may be violating Missouri consumer protection laws by misleading users with politically skewed outputs presented as objective information.
Bailey’s probe focuses on how the AI tools rate or assess public figures, citing a troubling example where OpenAI’s ChatGPT, Meta AI, and Google’s Gemini ranked President Trump last among recent presidents on the issue of antisemitism. Microsoft’s Copilot declined to respond. The attorney general called such responses “deeply misleading” and part of a broader pattern of politically manipulated content.
In formal letters sent this week, Bailey demanded records showing how the companies train their models, select inputs, and design algorithms that control chatbot responses. He emphasized the need for transparency to ensure companies are not using their platforms to push partisan propaganda.
“Missourians deserve the truth, not AI-generated propaganda masquerading as fact,” Bailey wrote.
Bailey also raised concerns that these responses could constitute deceptive practices under the Missouri Merchandising Practices Act. That law prohibits false advertising and consumer fraud. He stated that presenting biased or manipulated information—particularly when it affects public discourse and targets a political figure—may fall within the statute’s scope.
This inquiry follows growing national scrutiny of artificial intelligence and its potential for censorship. It also comes amid backlash faced by xAI, Elon Musk’s AI venture, after its chatbot Grok was found generating antisemitic responses. xAI later added restrictions and removed offensive outputs.
Bailey’s investigation reflects broader conservative concerns that major tech platforms disproportionately silence or misrepresent conservative viewpoints. His office intends to hold AI developers accountable for what he views as algorithmic censorship and deceptive content presented to millions of users.
