UN Warned by AI CEOs: Humanity Risks Losing Control of Artificial Intelligence

The CEOs of OpenAI and Anthropic warned the UN Security Council that humanity risks losing control over artificial intelligence, calling it an existential threat no single country can address alone. They urged international cooperation to prevent catastrophic outcomes. OpenAI's Sam Altman said AI systems could self-improve recursively, making decisions too complex for humans to understand or intervene in. He also warned against concentrating power in too few hands, proposing principles of human control, broad benefit, and individual empowerment. Anthropic's Dario Amodei cited risks of AI misuse, such as bioweapons creation, and loss of control over model capabilities. He proposed international agreements, verification systems, and global testing standards, pledging to slow development as needed for safety.