What happened
Anthropic CEO Dario Amodei used a long post on X to defend the company's "race to the top" strategy. The post was written as a reply to Nathan Calvin, chief strategy officer at the advocacy group Encode AI.
Nine safety decisions
Amodei listed nine decisions he said cost Anthropic commercially but made AI safer. One of them was holding back the Mythos model for four months on safety grounds.
Why it matters
The exchange touches on an ongoing question: how leading AI labs balance safety commitments against competitive pressure. Amodei's post signals that the company frames its safety choices as worth the commercial cost.
