Anthropic CEO urges US to block AI chip sales to China
Anthropic CEO framed AI chips as a vital resource which functions beyond its capacity as a commercial product
Anthropic CEO Dario Amodei is again urging the US government to block advanced AI chip sales to China, directly opposing Nvidia CEO Jensen Huang’s push to ease restrictions. Speaking on a podcast, Amodei argued that powerful AI chips represent a national security chokepoint and should not be treated as ordinary trade goods.
Amodei said the US should avoid building data centres in China or exporting high-end AI hardware. He described advanced AI systems as “essentially cognition, essentially intelligence,” stressing that such capabilities should not be placed in the hands of authoritarian governments.
The Anthropic CEO framed AI chips as a vital resource which functions beyond its capacity as a commercial product. He used the comparison of selling Nvidia's H200 chips to China to "selling nuclear weapons to North Korea" for showing the dangers which he believes result from US-China AI technology development.
Amodei expressed on the podcast his belief that the development of equally powerful AI systems by both nations would create a more unstable situation which nuclear deterrence fails to control. He maintained that direct confrontation between AI superpowers would lead to higher chances of miscalculation because each power believed it could succeed in battle.
He also raised concerns about offensive cyber dominance, suggesting that sufficiently advanced AI could make entire computer systems transparent to one side unless matched by equally strong defences.
The economic development from artificial intelligence will proceed at a quick pace, according to Amodei, but he believes that safeguarding political rights will present greater difficulties. He proposed that developing countries in Africa should receive AI infrastructure development while the United States should prevent the export of critical AI technologies to authoritarian governments.
-
Pentagon threatens to cut ties with Anthropic over AI safeguards dispute
-
Samsung Galaxy Unpacked 2026: What to expect on February 25
-
Can AI bully humans? Bot publicly criticises engineer after code rejection
-
Steve Jobs once called google over single shade of yellow: Here’s why
-
GTA 6 trailer hits 475m views as fans predict record-breaking launch
-
AI productivity trap: Why workers feel overloaded despite efficient tools
-
Meta to launch ‘name tag’ facial recognition for smart glasses this year
-
Agentic AI dating era: Bots are replacing humans in romance
