OpenAI has called on the U.S. government to take the lead in building international standards for artificial intelligence, according to the Financial Times. The company said countries should coordinate how they assess AI models’ capabilities and risks, and OpenAI CEO Sam Altman is set to present the proposal to the United Nations Security Council this week. OpenAI said an international framework could be built on existing national AI safety research networks, with common standards for evaluating models and communication channels between the U.S. and other countries to share information on newly emerging threats. The company also warned about risks tied to recursive self-improvement, or RSI, where more AI training work is carried out by other AI systems. OpenAI said that without proper safeguards, AI could iterate to a point where humans can no longer supervise it effectively, weakening human control over the technology’s development. The proposal comes as concern over AI safety has intensified. Anthropic and OpenAI have recently disclosed incidents involving AI models intruding into external systems, while Sam Altman, Anthropic CEO Dario Amodei, and SpaceX CEO Elon Musk earlier this month called for a slowdown in AI development. U.S. President Donald Trump has opposed slowing AI, saying, 「Whoever wins AI wins everything.」
OpenAI on Monday called on the U.S. government to lead the creation of international AI standards and said countries should coordinate how they evaluate AI models’ capabilities and risks, according to the Financial Times.
OpenAI CEO Sam Altman is scheduled to present the proposal to the United Nations Security Council this week. The company said existing networks of national AI safety research institutions could be used to build a common international standard for model evaluation. It also urged the U.S. to establish safety communication channels with other countries so newly emerging threats in AI can be shared across borders.
OpenAI flags recursive self-improvement risk
OpenAI specifically warned about risks tied to recursive self-improvement, or RSI, a scenario in which more AI model training work is carried out by other AI systems. The company said that without proper safety measures, AI could iterate quickly to a level that humans can no longer supervise effectively, causing people to gradually lose practical control over AI development.
Safety concerns remain in focus
The proposal comes as concern over AI safety has been rising. Anthropic and OpenAI have both recently disclosed incidents involving AI models intruding into external systems. Earlier this month, Sam Altman, Anthropic CEO Dario Amodei, and SpaceX CEO Elon Musk also called for a slowdown in AI development.
U.S. President Donald Trump has opposed slowing AI development and said, 「Whoever wins AI wins everything.」
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
