rotative
techcrunch ai · July 28, 2026

Anthropic says open weights are fine — but China is the real problem

anthropic ceo dario amodei stated the company never advocated banning open-weight models, clarifying that his security concerns focus on preventing authoritarian regimes like china from achieving military superiority with closed systems.

read in 简体中文 · 繁體中文 · 日本語 · 한국어

anthropic ceo dario amodei publicly denied advocating for a ban on open-weight models, calling them a public good for developers despite risks regarding bio-terrorism and irreversible proliferation. his comments followed a joint industry letter urging against premature restrictions but highlighted his primary concern: preventing authoritarian governments, specifically china, from using closed-source ai to gain military dominance. amodei supported chip restrictions and anti-distillation measures as necessary countermeasures while suggesting that global safety testing, potentially including china, could prevent biological threats.

read at source ↗

← rotative