President Trump has signed an executive order establishing a single national standard for regulating artificial intelligence, a move aimed at limiting the power of individual states to set their own rules. According to CNBC, the order is designed to create one federal framework rather than a patchwork of state-by-state laws.
The legal footing is already in question. NPR reports that the order attempts to preempt existing state AI laws through executive action, and that this approach "may not be legal" — a signal that court challenges could follow, since the authority to override state legislation by executive order is contested.
The response from industry and experts has been mixed to critical. According to gHacks, Anthropic is pushing back against a US order restricting its Claude Fable 5 model, a stance the outlet says is backed by cybersecurity experts. Meanwhile, Vox published a sharply critical assessment, arguing that Trump has "found the worst way to regulate AI."
Taken together, the sources describe a sweeping federal effort to centralize AI oversight in Washington — and an early wave of pushback spanning legal scholars, a major AI developer, and the tech press. Supporters of a unified standard typically argue it spares companies from navigating dozens of conflicting state rules, while critics warn that stripping states of authority could weaken protections and concentrate decision-making.
Why it matters: how AI is governed in the US — and whether that power sits with states or the federal government — will shape what these fast-moving technologies are allowed to do, and this order sets up a fight that courts may ultimately have to resolve.