The White House Is Making Up Its Rules for AI in Real Time
Anthropic still canโt distribute Claude Mythos or Fable 5 after running afoul of the Trump administration. But no one can say exactly what the company did wrong.
Anthropic still canโt distribute Claude Mythos or Fable 5 after running afoul of the Trump administration. But no one can say exactly what the company
Read Full Story at Wired โThe White Houseโs handling of AI regulations under the current administration signals a troubling shift toward opaque, ad hoc policymakingโone where rules emerge not from congressional debate or public notice, but from the instinctive reactions of a single administration. When Anthropicโs latest models were blocked without clear justification, the incident exposed a deeper vacuum in governance: America lacks a coherent framework for AI oversight, leaving companies to navigate a minefield of shifting political signals rather than stable legal standards. This isnโt just a corporate headache; itโs a systemic risk. Without predictable guardrails, innovation stalls, investment hesitates, and the U.S. cedes ground to competitors like China, where state-driven AI policy at least offers clarityโhowever heavy-handed. The backdrop matters here. The Trump administration has repeatedly signaled skepticism toward AI, framing it as a threat to jobs, security, or cultural norms without articulating a consistent alternative. But its approach isnโt an outlierโit reflects a broader erosion of institutional trust. Agencies like the Commerce Department, once seen as neutral arbiters, now operate under leaders who treat technology policy as an extension of partisan messaging. This creates a perverse dynamic: companies must curry favor with the White House to deploy cutting-edge tools, while the public gets no chance to weigh in on what risks are being prioritizedโor ignored. What happens next is anyoneโs guess. Will other firms quietly self-censor to avoid similar rejections, or will they push back with lawsuits challenging the administrationโs authority? Congress could step in, but gridlock makes that unlikely. More plausibly, weโll see a patchwork of state-level rules and industry self-regulation, deepening fragmentation. The bigger question is whether this moment forces a reckoning: Will the U.S. accept that AI governance canโt be left to improvisation, or will it double down on a system where power trumps process? The answer will shape not just the future of tech, but the balance between innovation and control in the 21st century.

