Anthropic CEO Dario Amodei publicly criticized the U.S. administration’s approval of chip sales to China and GPU manufacturers at the World Economic Forum in Davos on Tuesday.
The U.S. administration recently approved the sale of Nvidia’s H200 chips and an AMD chip line to authorized Chinese customers, reversing an earlier ban. These high-performance processors are used for AI applications. Amodei’s comments were notable as Nvidia is a major partner and investor in Anthropic.
Responding to a question about the new regulations, Amodei stated, “The CEOs of these companies say, ‘It’s the embargo on chips that’s holding us back.'” He predicted the decision would negatively affect the U.S. “We are many years ahead of China in terms of our ability to make chips,” he told Bloomberg‘s editor-in-chief, adding, “I think it would be a big mistake to ship these chips.”
Amodei emphasized the “incredible national security implications” of AI models, which represent “essentially cognition, that are essentially intelligence.” He envisioned future AI as a “country of geniuses in a data center,” illustrating this with the image of “100 million people smarter than any Nobel Prize winner,” controlled by a single nation.
He characterized the administration’s actions as “crazy.” Amodei drew a comparison, stating, “It’s a bit like selling nuclear weapons to North Korea and [bragging that] Boeing made the casings.”
Nvidia provides the GPUs powering Anthropic’s AI models, despite Anthropic utilizing Microsoft, Amazon, and Google servers. Nvidia announced an investment of up to $10 billion in Anthropic two months prior. This financial relationship included a “deep technology partnership” aimed at optimizing each other’s technology.
Anthropic has raised billions and is valued in the hundreds of billions. Its Claude coding assistant is recognized as a top-tier AI coding tool for complex real-world projects.
Anthropic’s stance may reflect concerns about Chinese AI development, aiming to draw Washington’s attention through stark comparisons. Amodei’s ability to vocalize such criticisms at Davos without apparent concern for business repercussions suggests an evolving landscape in AI leadership, where traditional constraints like investor relations and strategic partnerships may no longer apply.
