Trump Dismisses AI Extinction Warnings, Pledging to Keep the China Lead
President Trump said he has no concerns about AI leading to human extinction and that the US leads China in AI by roughly a year, publicly rejecting the risk estimates that Anthropic and OpenAI staff endorsed this week.
The political system’s answer to this month’s AI safety warnings arrived on September 11, and it was a flat rejection. Asked upon departing Dallas whether he has concerns about AI leading to human extinction, President Trump said, per Bloomberg, “No, I don’t have any.” CNBC’s coverage confirms he instead emphasized the race: the US leads China in AI by roughly a year, “we’re leading now… substantially,” and “we’re going to keep it that way.”
The Administration Position Is Now On the Record
The contrast with the past two weeks could not be sharper. Anthropic’s alignment science lead publicly endorsed a greater-than-10% chance of AI-caused human extinction this decade; its researcher Jacob Coxon resigned over labs “gambling with our lives”; OpenAI’s chief scientist published that no lab has solved alignment and its system card conceded degraded oversight of GPT-6. The President’s response to those warnings, delivered days later, is that the existential-risk framing matters less than the China race. That is not a fringe position; it is the operating assumption of the administration that drove the G20’s light-touch Carolina Principles and is reportedly dangling Nvidia chip access in diplomatic negotiations.
What This Splits, Practically
The US regulatory question now has a visible answer at the top: no slowdowns, no extinction framing, race-based policy only. That compresses the realistic paths for the ideas floating this month. A congressional safe harbor for coordinated lab slowdowns is unlikely to advance without White House support; the superintelligence ban bill was never going to pass this Congress; and state-level enforcement like the California AG’s OpenAI investigation becomes the main US check, alongside whatever the EU enforces. For developers and enterprises, the planning assumption for the US market is acceleration: faster model releases, lighter federal touch, and safety verification pushed onto vendors and states. For the labs, the political wind is at their backs, which makes the internal voices calling for restraint, on the record this month, the counterweight to watch as the Stargate-scale training runs continue.
Get Insanely Good at AI
The book for developers who want to understand how AI actually works. LLMs, prompt engineering, RAG, AI agents, and production systems.
Keep Reading
gr.Workflow Turns AI Pipelines Into Deployable APIs
Learn how to model, debug, expose, and deploy multi-step AI pipelines with Gradio Workflow and daggr.
G20 Unanimously Endorses the US Light-Touch Carolina Principles on AI
All G20 members, including China, endorsed the non-binding Carolina Principles for AI governance at the Chapel Hill ministerial on September 2, favoring sector-specific rules over comprehensive regulation.
Anthropic Names DeepSeek, Moonshot, and MiniMax in Industrial-Scale Claude Distillation Campaigns
Anthropic's September threat intelligence report details seven harm areas including industrial-scale distillation campaigns by DeepSeek, Moonshot AI, and MiniMax using roughly 24,000 fraudulent accounts to extract capabilities from over 16 million Claude exchanges.
Sanders and Casar Introduce a Bill to Ban Superintelligence, With 20-Year Prison Terms
Senators Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act on September 3, which would permanently prohibit AI matching or exceeding human cognitive performance and impose criminal penalties of up to 20 years.
Two-Week Gap Contradicts Kimi K3 Distillation Accusations
Frontier AI researchers dispute White House claims that Moonshot cloned Anthropic models, citing architectural differences and impossible training timelines.