Sam Altman addresses UN Security Council on AI safety

Summary

OpenAI CEO Sam Altman addressed the United Nations Security Council, arguing that powerful AI must remain under human control and that AI safety requires international cooperation. He said no level of catastrophic risk is acceptable, whether put at 10% or 0.1%, and that models should not be trained without a strong case for alignment, monitorability, and safety guarantees. He added that OpenAI has unilaterally slowed down in the past and will do so again. Press reports say Anthropic CEO Dario Amodei addressed the council the same day.

Why it matters for our work

Frontier lab leaders framing the debate around human control and global norms, rather than speed, is a signal for every organization adopting AI. The pace at which international safety norms settle should feed into adoption and development plans.

Translated from the Korean original. Summaries may be translated and edited. Commentary reflects our perspective; forecasts remain the source’s views.