Trump rejects UN AI pacts to build American AI Force
- Donald Trump rejected UN AI safety pacts and proposed a state-steered AI Force to outpace foreign rivals.
- Silicon Valley remains split as Anthropic demands binding global oversight while Nvidia opposes compute limits.
- Tech leaders called for development pauses as frontier labs face falling private valuations and open-weight competition.
Donald Trump just killed the global AI safety consensus. Speaking in New York, the former president rejected international oversight before the United Nations and announced plans for a state-directed American "AI Force" to accelerate military and industrial compute capacity.
On Sep 24, 2026, Trump rebranded artificial intelligence as superintelligence before UN delegates, declaring that American development will proceed without permission under internal monitoring by the Department of Justice. White House advisor Michael Kratios reinforced the sovereign stance on The AI Daily Brief, arguing that national growth and liberty face far greater risk from international regulators than from rapid technology deployment.
A coalition of 20 nations led by Finland pushed for mandatory international safety testing and a binding global verification body. Neither the United States nor China endorsed the plan, leaving global governance dead on arrival while superpowers race for hardware dominance.
The pivot follows weeks of maneuvering across Silicon Valley. On Sep 21, 2026, tech leaders including Sam Altman, Dario Amodei, and Elon Musk raised eyebrows by simultaneously calling for pauses on AI deployment. Host Marty Bent noted on TFTC that the sudden corporate panic looked like an orchestrated move to shield frontier labs from legal liability as secondary equity valuations plummeted.
"Bitcoin does not care about your macro headlines."
- Marty Bent, TFTC: A Bitcoin Podcast
While financial markets and state regulators grapple with technological dislocation, frontier labs are hitting commercial walls. Corporate clients are abandoning proprietary models for self-hosted open-weight alternatives to avoid intellectual property risks, leaving tech giants eager to gatekeep high-margin enterprise tools instead of selling raw tokens.
Trump's answer to the commercial slowdown is pure state direction. Speaking on Bitcoin And on Sep 21, 2026, host David Bennett dissected Trump's plan for a federal AI Force overseen by an executive czar. Bennett pointed out that White House officials offered zero clarity on whether the unit would operate as a military branch, a cabinet department, or a civilian bureaucracy, despite targeting high-IQ applicants.
The policy shift exacerbates an escalating feud among technology executives. At the UN Security Council, Anthropic chief executive Dario Amodei pleaded for binding international treaties and mandatory model testing to avert catastrophic misuse. Conversely, Nvidia chief executive Jensen Huang has aggressively lobbied against hardware restrictions that threaten chip sales.
The era of voluntary safety pledges is officially over. By replacing international treaties with state-steered expansion, Washington is treating compute capacity as national firepower. The race for AI dominance will not be governed by international consensus, but by who controls the power grid and the silicon supply.