OpenAI urges US to set global safety standards for AI
What's the story
OpenAI has called on the United States to take the lead in creating international safety standards for advanced artificial intelligence (AI) systems. The appeal comes as world leaders gather for the United Nations General Assembly, where discussions on AI's safety, competition, national security, and self-improvement capabilities are gaining prominence.
Safety standards
Framework for AI safety
OpenAI has proposed that the US should lead the way in creating a global safety framework for advanced AI.
The company suggested this framework should align national regulations with international standards and include technical measures for frontier AI.
Their proposals also include concepts like recursive self-improvement (the ability of an AI system to improve its own capabilities) and shared methods to measure risks and report serious incidents.
Regulatory concerns
Incompatible approaches to AI management
OpenAI has warned that without greater coordination, countries could adopt incompatible approaches to managing advanced AI.
The company believes the US has an opportunity to shape the emerging international system rather than letting fragmented rules develop independently.
This proposal comes as governments grapple with the challenge of balancing technological progress with concerns over systems that could outpace existing safeguards.
UN address
UN Security Council meeting on AI
OpenAI CEO Sam Altman will address the UN Security Council on Wednesday, focusing on artificial intelligence and international security.
The meeting was called by France, which currently holds the presidency of the council for September.
This marks a major step in global discussions about AI's potential risks and benefits, as world leaders consider how best to regulate this rapidly evolving technology.
Development pace
Pacing approach to AI development
OpenAI has proposed a "pacing" approach to AI development, emphasizing that research into aligning AI systems with human intentions should keep pace with advances in capability.
The company also stressed the need for international coordination to establish common benchmarks and procedures for identifying and responding to risks as capabilities evolve.
This comes amid strategic competition between the US and China over AI technology.