Anthropic says it does not support banning open-weight AI models, while warning about authoritarian AI risks

Anthropic CEO Dario Amodei says the company has never advocated for a ban on open-weights models, including those from China. He argues non-dangerous open models are a public good, but says his main concern is authoritarian governments building more powerful AI systems and using them for military dominance or repression.

The post positions Anthropic against protectionist bans while keeping pressure on governments to address frontier AI security risks. It reflects a broader split in AI policy between promoting open innovation and limiting the strategic advantages that advanced models could give authoritarian states.