OpenAI is asking the U.S. government to lead an effort with other countries to develop global safety and security standards for building cutting-edge artificial-intelligence systems.
In a proposal published Monday, the company suggested that countries worldwide should set common measures for evaluating forthcoming self-improving AI systems and establish secure channels to share emerging threats.
The U.S. Center for AI Standards and Innovation could work with the emerging network of AI safety institutes around the world-such as existing groups in the U.K., France, and Singapore-to create the standards, the company said.
The call to action comes as global leaders convene in New York City this week for the United Nations General Assembly and as President Trump prepares to host Chinese leader Xi Jinping in Washington. Earlier this month, top tech leaders called for industrywide coordination to slow down future AI development.
This summer, a collective of agents that OpenAI was testing internally gained internet access and hacked the AI platform Hugging Face, setting off an industry reckoning with the cybersecurity risks of advanced AI models.
The company said risks similar to the Hugging Face incident could become more severe and common following the onset of self-improving AI systems. OpenAI recently learned it was the latest target of an AI-powered intrusion. Last week, a team of independent security researchers revealed they used Anthropic's Claude software to gain access to an OpenAI employee's ChatGPT account, giving them a way to read the company's private code base.
OpenAI's chief scientist Jakub Pachocki recommended this month that international coordination on AI should become a priority for governments around the world to address the risks of self-improving AI systems. Anthropic Chief Executive Dario Amodei last month offered a proposal to slow down more advanced AI development, and asked Washington to issue a narrow waiver for tech companies to discuss safety and security measures. The Trump administration has resisted the calls. Treasury Secretary Scott Bessent said last week that AI labs shouldn't receive a "liability exemption, which is what they are asking for." Bessent said that "the best way to guarantee is that the creators are liable for what they build and generate."
President Trump chimed in on social media Monday: "Whoever wins AI, WINS! We are leading now over China, and everyone else, and I'm going to keep it that way! I'm not going to stifle Growth, of something that will be bigger than the Industrial Revolution, or the Internet, itself."
He added that the Justice Department and other law enforcement bodies could "rein things in if we have to."