Missouri Attorney General Andrew Bailey has launched an investigation into Google, Microsoft, OpenAI, and Meta over their AI chatbots' responses to a question about ranking the last five presidents on antisemitism. The probe alleges "deceptive business practices" because the chatbots ranked Donald Trump unfavorably. Bailey claims the AI-generated opinions are misleading to Missouri consumers and has demanded internal documents on AI training and weighting from the companies.
The investigation is based on a conservative group's test that posed the ranking question to six chatbots, including the four mentioned above, plus X's Grok and the Chinese LLM DeepSeek. Both Grok and DeepSeek ranked Trump first. However, Bailey's investigation contains basic factual errors, such as claiming Microsoft's Copilot ranked Trump last when it actually refused to answer.
Bailey's letters to the companies accuse them of making "factually inaccurate" claims and demand explanations for why their chatbots produced results that appear to disregard objective historical facts. Critics argue that the investigation is an abuse of power and a "fishing expedition" intended to chill speech. The probe raises questions about the role of AI in political discourse and the potential for government control over AI opinions.
The investigation has sparked controversy, with many questioning Bailey's motives and the validity of his claims. Legal experts say that the government's attempt to punish companies for expressing opinions it dislikes is a clear violation of First Amendment principles. The outcome of this investigation remains to be seen, but it has already sparked a heated debate about the role of AI in society and the limits of government intervention.
