Trump's April 15 comments on AI and banking reveal the core tension in his administration's approach to artificial intelligence: a genuine desire to position the U.S. as a global AI leader competing against China, balanced against limited appetite for prescriptive regulation. The timing is significant—coming days after Anthropic released its Mythos model to a restricted group, with Treasury and Fed officials convening banks to discuss cybersecurity implications. Trump's willingness to say "there should be" government safeguards on AI suggests some openness to guardrails, a modest shift from his broader deregulatory posture.

The core disagreement between left and right hinges on timing and scope of intervention. The left argues that AI systems in banking require exhaustive pre-deployment safety testing, transparent model documentation, and clear liability frameworks before deployment. They point to the left-brain AI that can identify thousands of zero-day vulnerabilities in operating systems as proof that frontier AI systems carry unknowable risks. The right contends that perfect safety is impossible to achieve and that insisting on it amounts to a de facto moratorium. Trump's framework essentially trusts companies to deploy AI while maintaining government ability to intervene through mechanisms like a kill switch.

What's underreported is the internal contradiction in Trump's approach: his administration labeled Anthropic a national security supply-chain risk and directed the Pentagon to stop using its technology, yet simultaneously told banks to adopt that same technology for cybersecurity purposes. A federal judge in California issued a preliminary injunction blocking the supply chain designation, noting that "nothing in the governing statute supports the Orwellian notion that an American company may be branded a potential adversary," while an appeals court in Washington, D.C., denied Anthropic's request to temporarily halt the blacklisting. The net effect: Anthropic is excluded from DoD contracts but can continue working with other government agencies, and it is into that gap, excluded from the Pentagon but not from the Treasury or the Fed, that Bessent and Powell stepped. This schism reflects not policy coherence but institutional competition—the Pentagon vs. Treasury/Fed—and suggests the administration has not fully resolved its stance on which companies and which AI capabilities it trusts.

The left employs alarm-laden language about "demolishing" protections and "vacuums of accountability," while the right uses optimization language like "balance," "pragmatic," and "unleash innovation." The left treats Trump's banking AI endorsement as reckless; the right treats it as a measured acknowledgment of both promise and risk.