China Slams Anthropic CEO’s AI Slowdown Plan as “Cold War Playbook”

China’s state-backed Global Times has called China’s AI Cold War the real frame behind Anthropic CEO Dario Amodei’s safety essay. The publication branded Amodei’s three-step pacing plan a ‘Cold War playbook.’

Ad

Ad

Beijing Says Safety Talk Is Containment, Not Governance

The Global Times commentary said Amodei’s essay ‘may look like a rational statement focused on safety risks’ but is, in essence, a bid to ‘curb China’s AI development through technological barriers and regulatory monopolies.’

A follow-up piece quoted Xiang Ligang, a veteran tech policy observer, who called the rhetoric ‘inappropriate, hostile, and groundless.’

Tsinghua University’s Xiao Qian went further. He tied Amodei’s stance directly to commercial pressure from cheap Chinese open-source models. The White House’s own narrative lines up here.

It already flagged alleged model distillation by Moonshot AI, a claim detailed in the White House distillation case against Moonshot AI.

Amodei’s essay specifically called on Washington to crack down on distillation and tighten security around model weights. That is the same policy ask.

The reaction split the U.S. industry. Sam Altman said OpenAI would match Anthropic’s pledge to allow third-party evaluators inside its labs.

Elon Musk posted three words: ‘Dario is right.’ Yet NVIDIA, Meta, and Microsoft warned against open-weight AI restrictions, a direct counterweight to Amodei’s containment line.

Trump, speaking to reporters in Doonbeg, Ireland on Sunday, rejected any slowdown outright. ‘Whoever wins AI wins,’ he said, per Reuters.

The Trump administration has been weighing restrictions on Chinese AI models. But Beijing is refusing to frame the standoff as a shared safety project.

China’s MFA posted on X that AI needs an ‘open, inclusive’ approach. It urged parties to drop ‘fear-mongering and confrontation.’ A Trump–Xi dialogue window around September 24 is the next pressure point.

Ad

Ad

Investor Implications: IPO Risk, Compute Scarcity, and the China Discount

The clash is not just diplomatic. It is repricing assets. Anthropic is in the IPO pipeline, with the Anthropic vs. OpenAI pre-IPO trade already live on markets, and Coinbase launching Anthropic pre-IPO futures after its SpaceX success.

Valuation talk has ranged as high as $2 trillion, rivalling SpaceX in the IPO market-cap race.

Tighter chip and compute policy is a direct tailwind for U.S. closed-model labs. Former White House crypto czar David Sacks warned the U.S. could lose the AI race, a frame Trump now echoes publicly.

The U.S. government has already moved, clearing Anthropic’s Mythos rollout to selected organizations, treating the lab as strategic infrastructure. That adds a policy premium to U.S. frontier models, and a scarcity signal for compute.

On the crypto side, Anthropic already holds a 20-year, ~$19B contracted lease through its TeraWulf data-center deal.

Any ‘win AI at all costs’ push from Washington is a demand signal for power and GPUs, the same stack Arthur Hayes frames as compute as the scarce asset in an agentic economy.

Meanwhile, DeepSeek is eyeing its own IPO to rival OpenAI and Anthropic, a sign Beijing will not self-slow. Investors should treat the China’s AI Cold War label as a policy tell, not a headline.

Until Washington, Beijing, or the labs flinch, capital will keep paying for compute and closed U.S. models, while pricing in a geopolitical discount on any AI name that needs China as a customer or governance partner.

From on-chain data to charting, explore these free crypto tools every investor should know.