Anthropic: No Open-Weights Ban, Curb China Chips
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic published an official position through CEO Dario Amodei stating the company has never backed an open-weights model ban
- Amodei's paper calls for restricting top chip sales to China, cracking down on Chinese distillation, and implementing global model safety testing
- OpenAI and Google are backing open weights per The New Stack, creating an industry split with Anthropic's call for testing over unrestricted release
- The Register frames the debate around Jensen Huang reportedly putting 'his thumb on the scales against open-weights fearmongering'
- David Sacks on X accused Anthropic of 'breathtaking' hypocrisy, arguing it 'maintains that it is entitled to train for free on all the world's output' while labeling competitor training on Anthropic's output as 'IP theft'
- Nathan Lambert on X called Amodei's piece 'reasonable' while noting 'nothing new in it' and that 'banning distillation is still dumb'
Why it matters: Anthropic's stance creates a tripartite industry split heading into Congressional debates over AI policy—Anthropic favors targeted China restrictions and testing, while OpenAI and Google back open weights. The position also drew pointed IP criticism from David Sacks on X, exposing tensions over fair-use training practices that cut across the entire sector.
