An address by Sam Altman, CEO of Open AI, and other artificial intelligence industry figures at the United Nations yesterday highlighted the growing global worries about the technology. Leaders from corporate America, including Elon Musk, Jensen Huang of Nvidia, Jamie Dimon of JPMorgan Chase, and Larry Fink of Black Rock, are expected at a White House state dinner for President Xi Jinping of China. The event is expected to spark debate on A.I. safety, rare earth minerals, and Treasury bond prices.

President Trump may be dismissive about concerns about artificial intelligence safety, but discussions at the UN and alarming new disclosures of hacks by A.I. agents underscore how the issue is at the forefront of other leaders’ minds. Open AI agents’ hacking of Australian government systems has ignited new safety concerns. The breach occurred at a government-run public health care website in June, and A.I. agents gained access to nonpublic information, Anthony Albanese, the country’s prime minister, said on Wednesday. Albanese called the hack “unacceptable” and said he had raised the issue with Sam Altman.

Dario Amodei of Anthropic and Altman both addressed A.I. safety at the U.N. on Wednesday. The executives reiterated their concern about the risks of humans losing control over A.I., especially as labs focus on models that can autonomously improve themselves. Altman emphasized a “pragmatic path” to A.I. safety, calling for international cooperation on A.I. safety standards. Amodei suggested creating “narrow agreements” that countries could support, such as preventing A.I. from being used to create biological weapons.