UN Security Council Convenes AI Safety Briefing With OpenAI, Anthropic, and DeepSeek Executives

Key Highlights

• Leaders from OpenAI, Anthropic, and China’s DeepSeek have received invitations to present before the UN Security Council regarding artificial intelligence safety concerns this Wednesday
• Dario Amodei, CEO of Anthropic, will participate via remote connection, with OpenAI’s Sam Altman also scheduled to speak
• DeepSeek’s founder Liang Wenfeng likely won’t be present, although participation plans remain subject to modification
• Amodei from Anthropic has advocated for a synchronized deceleration in cutting-edge AI advancement, a position endorsed by Elon Musk
• Treasury Secretary Scott Bessent announced that high-ranking American and Chinese representatives have committed to continuing artificial intelligence safety conversations before an anticipated Trump-Xi summit

On Wednesday, the UN Security Council will convene a session focused on artificial intelligence, featuring presentations from several prominent figures leading major AI companies. This gathering coincides with the UN General Assembly, bringing world leaders together in New York.

Sam Altman, chief executive of OpenAI, and Dario Amodei, who leads Anthropic, are confirmed participants in the proceedings. Amodei’s contribution will come through virtual attendance. Additional presenters include Yoshua Bengio, an AI researcher serving as co-chair of the UN’s artificial intelligence panel, along with Clement Delangue, CEO of Hugging Face.

DeepSeek, a Chinese artificial intelligence venture, and Moonshot, another startup, have received invitations to deliver remarks, per Reuters reporting. While Liang Wenfeng, DeepSeek’s founder, isn’t anticipated to appear physically, scheduling adjustments remain possible.

Attempts to reach DeepSeek and Moonshot for comment went unanswered.

Disagreement Emerges Over AI Development Pace

This gathering occurs amid mounting tensions regarding the appropriate velocity for artificial intelligence progress. Amodei of Anthropic has proposed a collective reduction in the speed of frontier AI advancement. Musk has publicly endorsed this perspective.

Contrasting viewpoints have emerged from Mark Zuckerberg at Meta and Jensen Huang at Nvidia, both contending that decelerating development is unwarranted.

Last week, UN Secretary-General Antonio Guterres cautioned global leaders that the rapid trajectory of AI innovation presents risks requiring immediate attention. His stance creates tension with US President Donald Trump’s assertion that existing safeguards are sufficient.

Washington and Beijing Commit to Enhanced AI Safety Dialogue

In related developments, Treasury Secretary Scott Bessent revealed Monday that top American and Chinese representatives have agreed to advance additional artificial intelligence safety conversations. This commitment precedes the anticipated meeting between Trump and Chinese President Xi Jinping in Washington scheduled for later this week.

State-controlled Chinese media outlets have criticized proposals to decelerate AI progress, characterizing such efforts as strategies to consolidate American technological superiority while excluding China from international AI governance frameworks.

Chinese artificial intelligence industry leaders have maintained relative silence regarding safety issues, including potential scenarios involving AI systems that might deceive users, circumvent supervision, or generate extensive damage.

The Security Council’s initial examination of AI-related risks occurred in 2023. During that gathering, Chinese representatives cautioned against allowing AI to become a “runaway horse,” while American delegates warned against employing the technology for suppression or censorship purposes.

This Wednesday’s proceedings represent the council’s most significant AI-focused meeting since those preliminary conversations, marking the first occasion where representatives from opposing perspectives in the development debate converge at this diplomatic level.