Anthropic CEO Proposes UN-Style Global AI Governance Body

Dario Amodei, the CEO of Anthropic, published a series of essays outlining a vision for artificial intelligence that balances human flourishing against risks of geopolitical instability and catastrophic biological threats. He proposes the establishment of a United Nations-style AI standards body to manage global oversight.

Amodei advocates for a verifiable international agreement with China, comparing the proposed framework to the Paris Climate Deal. He argues that joint governance by democratically elected governments is necessary to ensure safety, stating that he is concerned a single government or company could abuse the technology.

Critics of the proposal claim Amodei is influenced by a narrow circle of liberal policy experts, establishment academics, and figures from the Effective Altruism movement. These critics argue that such an ideological lean creates a blind spot regarding the actual reliability of international agreements with adversarial regimes.