OpenAI Calls on Washington to Set Worldwide AI Standards
U.S. Should Drive Global AI-Safety Rules
U.S. Should Drive Global AI-Safety Rules
Key Takeaway:
OpenAI is calling for U.S. leadership: The company wants Washington to spearhead international efforts to create technical standards for frontier AI, including systems that may eventually improve themselves.
It supports globally consistent rules: OpenAI is advocating shared benchmarks, coordinated national and international standards, and common procedures for reporting serious AI incidents.
Safety research should keep pace with AI progress: Rather than demanding an arbitrary halt, OpenAI promotes “pacing”—ensuring that alignment research and safeguards develop at least as quickly as AI capabilities.
The matter is expected to be discussed at the UN Security Council: CEO Sam Altman is set to brief the council on the implications of AI for international security, building on previous UN discussions about related risks.
Washington remains split over the approach: OpenAI favors greater international cooperation, while the Trump administration has opposed broad pauses in AI development, warning that slowing U.S. innovation could strengthen China’s strategic position.
OpenAI is urging the U.S. government to work with other countries in the establishment of global safety and security standards for advanced artificial-intelligence systems, on the grounds that international cooperation will become all the more important as AI becomes more powerful and possibly autonomous.
Also Read:
The AI giant, in a proposal released on Monday, advised that countries create common procedures for testing and assessing future self-improving AI systems and called on governments to set up secure channels through which they can share information on emerging threats. The company proposed that the U.S. Center for AI Standards and Innovation should cooperate with AI safety organizations in countries such as the United Kingdom, France, and Singapore.
The proposal is being made at the same time that world leaders are gathering in New York for the United Nations General Assembly and that President Trump is preparing to meet Chinese President Xi Jinping. It also shows the increasing worries within the technology industry concerning the cybersecurity risks linked to more advanced AI, especially in the light of incidents involving unauthorized access to online platforms and private accounts.
OpenAI has cautions that breaches like the most recent one at Hugging Face might become both more common and more severe as self-improving AI systems are developed. Jakub Pachocki, OpenAI's chief scientist, has asked that governments make international cooperation a top priority. Dario Amodei, CEO of Anthropic, has likewise requested that the development of more advanced AI be slowed down and that companies be allowed to discuss safety measures under some legal protection.
The Trump administration has opposed imposing wide-ranging restrictions on AI development, with Treasury Secretary Scott Bessent stating that companies should stay legally responsible for the systems they build and the President asserting that the United States needs to keep its leadership in the field of AI and must avoid any policies that might hinder technological and economic growth.
Ali Merchant is a seasoned financial markets professional specializing in technical analysis, treasury & capital markets, trading, research, and fund management. He is Strategic Advisor at Musaffa and Founder of TWT Learning, which provides education, research, and advisory services to fund and hedge managers and family offices. Ali has traded FX, FX options, US stocks & options, indices, commodities, oil and metals futures, and holds a CMT charter with memberships in AAPTA and CMT, US & Canada. He has worked with ABN AMRO, Thomson Reuters/Refinitiv, MAK Allen & Day Capital, and Bridge Information Systems, produced TA reports for Bloomberg and Reuters, and trained 2,000+ market participants across North America, the GCC and Asia.