Humans First urged the White House to require testing and approval of advanced AI models before release.

Humans First, a conservative group whose mission is ensuring “technology should serve humans … not replace them”, sent a letter urging the administration to require government testing and approval before the release of advanced AI models. More than 60 noted political figures and activists, including Steve Bannon, signed the letter.

The letter calls for a mandatory federal process to review AI systems prior to deployment. It states that companies should not release high-capability models without demonstrating safety through testing. Humans First warns that military, nuclear, and financial systems are exposed because AI models can rapidly identify cyber vulnerabilities and develop exploitation tactics. The group calls for safety checks similar to those used in aviation and nuclear systems. Further, the letter voices concerns that “No private corporation should have unilateral authority to deploy technologies that could profoundly impact the national security and stability of the United States without meaningful evaluation and safeguards.” 

The stance highlights a split among Trump supporters about how to approach AI. The White House’s approach to date has been laissez-faire, limiting regulations in favor of allowing AI companies to develop rapidly with minimal guardrails. Recently, this approach shifted to a more guarded approach with the release of Anthropic’s Mythos, but still falls short of regulating frontier AI model releases. 

The coalition also raises concerns about job losses to AI and its economic impact.