A statement from Dario Amodei on Anthropic’s commitment to American AI leadership

Source: Anthropic Date: 2025-10-21 URL: https://www.anthropic.com/news/statement-dario-amodei-american-ai-leadership

Summary

Dario Amodei published a statement affirming Anthropic’s alignment with the Trump administration on AI policy — citing the $200M DOD agreement, government-wide Claude access, and support for the AI Action Plan. Defended against political bias criticism by citing a study showing Claude Sonnet 3.5 as “less politically biased than models from most other major providers.” Claimed Anthropic is “the only frontier AI company” restricting chip sales to Chinese entities. Advocated for federal AI standards over state-level regulation (implicitly pushing back on California’s regulatory approach).

Implications

• Government/defense / political positioning thread. October 2025 — Amodei is explicitly aligning with a Republican administration, which is a departure from the typical Silicon Valley posture. The political bias defense and DOD partnership framing suggest Anthropic was facing pressure from conservative critics and government customers simultaneously.
• “Only frontier AI company” restricting China chip sales. This is an extraordinary claim — if accurate, it means OpenAI, Google, and Meta do not voluntarily restrict chip sales to Chinese entities, while Anthropic does. The claim may refer to financial investments or partnerships rather than hardware sales; the specifics matter.
• Federal over state standards. Pushing for federal AI standards while implicitly opposing state-level regulation (SB 53 support noted as moderate) is the position that neutralizes California’s EU-AI-Act-style approach — which Anthropic has publicly supported in other contexts. The tension is real.
• Political bias study citation. A study showing Claude Sonnet 3.5 is less biased than “most other major providers” — but not all — is carefully worded. This became the foundation for the November 2025 political even-handedness evaluation methodology.
• Watch: whether the DOW supply chain risk designation (March 2026) came despite this October 2025 alignment statement, suggesting the conflict was about specific capability restrictions rather than political posture.