Sam Altman, OpenAI’s chief executive, plans to brief the UN Security Council on Wednesday as it considers the risks artificial intelligence may pose to international security. He is expected to be joined in the session by senior representatives of Anthropic, as well as statements from Chinese artificial intelligence firms including DeepSeek and Moonshot.
The meeting has been called by France, which holds the rotating presidency of the 15-member council this month. It comes as world leaders gather in New York for the annual UN General Assembly, putting several of the companies building the most advanced AI systems before the UN’s main security body.
DeepSeek will be among the companies invited to address the council. Its founder, Liang Wenfeng, does not plan to attend, and it is unclear who will represent the Hangzhou-based company. Moonshot has also been invited, though it is unknown whether it will send representatives. DeepSeek and Moonshot did not respond to requests for comment.
Plans for the briefing could still change at the last minute. The format of the session, its duration, whether it will be public or closed-door, and the positions to be set out by individual companies have not been specified.
France’s concept note for the meeting points to the prospect of AI being used maliciously and to the risk that people could lose control of advanced models. Several industry leaders have called for a coordinated slowdown in the development of increasingly powerful systems. They have warned that such systems could soon improve on their own and slip beyond human control.
The council has previously considered the subject. During a meeting on AI risks in 2023, China warned that artificial intelligence could become a “runaway horse”. The U.S. raised concerns over the technology’s possible use for suppression or censorship.
DeepSeek’s participation follows its sudden emergence in January 2025, when its R1 reasoning model attracted worldwide attention. The model’s low cost had a damaging effect on US technology stocks, putting the Hangzhou company at the centre of debate over the pace and cost of Chinese AI development.
Wednesday’s session will bring DeepSeek and Moonshot into the same diplomatic forum as OpenAI and Anthropic. The companies’ specific statements remain unknown, though the meeting will give the Security Council a chance to hear directly from firms whose systems are central to the debate over AI misuse, international security and human control of advanced models.
Comments (0)