DeepSeek to Brief UN Security Council on AI Risks
nashnova research
Chinese AI firm DeepSeek will address the UN Security Council on AI risks this week, appearing alongside OpenAI and Anthropic — marking the first time Chinese AI companies have stepped into the highest-level multilateral security forum, as Washington and Beijing remain deeply divided on how to govern the technology.
What is this Security Council meeting about?
The UN Security Council is set to hold a dedicated session Wednesday on AI and international security.
Weeks before the meeting, leading AI executives co-signed a call to coordinate a slowdown of increasingly powerful AI systems, warning they could soon self-improve and slip beyond human control.
This means → the session is not a tech showcase — it is a security-alarm-driven political meeting asking one core question: who should govern the risk of AI going out of control.
Who will be there — and who won't?
OpenAI CEO Sam Altman plans to attend and brief the council; a senior Anthropic representative is also expected.
On the Chinese side, DeepSeek and Moonshot have been invited to deliver statements — but DeepSeek founder Liang Wenfeng will not attend in person, and arrangements remain fluid.
In plain terms = Chinese AI firms are willing to send a message, but the founder staying away keeps one foot in and one foot out.
How wide is the US-China divide on AI governance?
UN Secretary-General Guterres warned last week that the risks of rapid AI development cannot be ignored; President Trump countered that existing safeguards are sufficient and that questioning AI progress benefits China.
Chinese state media framed calls to slow AI development as a "Cold War playbook" — designed to lock in Washington's lead in frontier technology and shut Beijing out of global AI governance.
This reflects a mirror-image disagreement: the US worries Chinese AI is opaque; China worries the rules are designed to freeze its progress.
Is there any movement on the bilateral track?
Treasury Secretary Scott Bessent said Monday that the US and China have agreed to hold further talks on AI safety, following preparatory meetings ahead of this week's Trump–Xi summit in Washington.
This means → the multilateral track (Security Council) and the bilateral track (direct US-China talks) are running in the same week — but whether either produces real consensus remains an open question.
Why does this meeting matter?
This is the first time Chinese AI companies have appeared publicly in a multilateral security framework. When the Security Council first discussed AI risks in 2023, China warned AI must not become an "unbridled horse"; the US warned AI must not be used for censorship or repression.
In plain terms = the two sides were talking past each other two years ago, and that dynamic has not fundamentally changed — but this time Chinese companies are speaking directly, which signals at least one shift: the AI governance conversation is expanding from governments to firms.
市场有风险，内容仅供研究参考，不构成投资建议。