Anthropic, Google, OpenAI, and xAI recently secured Defense Department contracts worth up to $200 million. A planned executive order could put those funds in jeopardy if they don’t fall in line.
UPDATE 7/23: As part of an “AI Action Plan” unveiled this week, one of the executive orders does indeed target “woke AI in the federal government.”
The EO essentially continues the Trump administration’s campaign against diversity, equity, and inclusion (DEI), and calls out Google for the historically inaccurate images its AI created last year. (No mention of Grok’s MechaHitler tirade, Trump’s AI-generated pope image posted on the official White House X feed, or the AI video Trump shared depicting President Obama getting arrested in the White House.)
The EO argues that the federal government “should be hesitant to regulate the functionality of AI models”, but also says it should not allocate funds to “models that sacrifice truthfulness and accuracy to ideological agendas.”
The EO says agency heads should only use large language models (LLMs) developed in accordance with two principles: truth-seeking and ideological neutrality. Trump is giving several agency heads—the director of the Office of Management and Budget, the Administrator for Federal Procurement Policy, the Administrator of General Services, and the Director of the Office of Science and Technology Policy—120 days to figure out how to assess that.
