OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei addressed the United Nations Security Council in separate briefings focused on international AI safety and governance. The high-level addresses follow recent warnings from Anthropic researchers regarding existential risks posed by advanced AI systems.

Altman urged governments to establish democratic oversight over AI development to avoid centralized control by San Francisco laboratories alone. He outlined a balance between innovation and safety based on human control, widespread public benefit, and individual empowerment, while cautioning against excessive “doomerism.”

Amodei advocated for formal international agreements, including global bans on using AI for biological weapons and standardized testing protocols. Emphasizing Anthropic’s focus on safety, Amodei reiterated that mismanaged frontier models could pose systemic risks to humanity.

Why it matters

  • Signals rising pressure for unified international regulations, which could impact frontier model deployment rules and safety compliance costs.

  • Highlights the divergence between OpenAI’s economic-focused pitch and Anthropic’s existential risk framework in global policy settings.

Source: theguardian.com