DeepSeek and other Chinese AI firms, including Moonshot, have been invited to brief the UN Security Council on AI safety this week, according to sources. The same institutions that help organize global force are now hosting a discussion about the machines that could shape it, with invited firms expected to make statements.
Who Gets to Speak
The Reuters report on the invitation was published on Sep. 22, 2026, at 09:18:16 GMT. The briefing is set as a UN Security Council discussion on AI safety and governance, and the invited firms are expected to make statements. That means the people building the systems are being brought into the room where states talk about how to manage them. Ordinary people, meanwhile, get the fallout.
The Diplomat published a separate analysis on Sep. 22, 2026, at 15:00 GMT, arguing that America's AI safety dilemma is tied to the Trump-Xi Summit and to the broader competition between the U.S. and China over AI technology and policy. The whole thing reads like a contest between rival power blocs, each claiming to protect the public while jockeying for control over chips, compute, and access to American models.
That analysis said the Trump administration faces pressure at home to slow AI development, while also fearing that any slowdown would help China. So the answer from above is not restraint, but more control. It said the administration sees tighter controls on chips, compute and access to American models as a possible way to slow Chinese AI instead. The summit could shape policy on export controls and diplomatic signaling. The people are told this is safety. The machinery says otherwise.
Who Pays for the Race
The Diplomat piece said Anthropic CEO Dario Amodei published an essay on Sept. 12 titled “We Must Pace the Frontier,” arguing that leading labs should deliberately slow how fast their models gain new capabilities. It said Sam Altman said OpenAI would match Anthropic’s commitment to host independent evaluators, and Elon Musk replied, “Dario is right.” Even here, the fix comes from the same elite circle that built the problem: labs, CEOs, evaluators, and the state apparatus circling the same table.
The piece said U.S. President Donald Trump later called fears of an AI catastrophe a hoax and said that amid the backlash against AI and data centers, “the only one that is happy about it is China.” That’s the language of a ruler trying to turn public fear into a nationalist script. It doesn’t answer the danger. It just picks a rival.
The analysis said Xi Jinping’s arrival in Washington this week is the first state visit by a Chinese leader in more than a decade. It said Trump said he wants to stay ahead of China because “whoever wins AI wins,” and that Treasury Secretary Scott Bessent warned there is “no day after tomorrow if China wins” on AI and that “nothing [else] would matter.” The rhetoric is pure domination. Win AI, win everything. Lose AI, lose the future. The rest of society gets reduced to collateral in a race run by states and firms.
The piece said a new Politico poll found 63 percent of respondents saw a real chance that AI could wipe out humanity, while a CBS News/YouGov survey found most Americans want AI development slowed, though not stopped. It said Politico found Americans split, 43 percent to 37 percent, on whether falling behind other countries is acceptable in exchange for safer AI, and that Trump’s 2024 voters preferred staying in the AI race, 50 percent to 38 percent, over prioritizing safety issues. Public fear is plain. The policy response is still a contest over who gets to keep building.
What the Power Brokers Call Safety
The analysis said Amodei suggested an “attempt to coordinate with authoritarian governments, to the extent this is possible, while taking seriously the challenges of verifying compliance,” and said Beijing is at the top of the list. He also urged keeping “powerful AI chips or semiconductor manufacturing equipment” out of Chinese hands, closing “chip smuggling operations” and remote access to data centers, cracking down on unauthorized distillation and hardening labs against theft of model weights. That’s not public control. It’s a managed choke point, designed by insiders and enforced through the same security logic that always protects the powerful first.
Amodei agreed with Bessent that “a Chinese lead in AI would pose grave danger for the United States and the world.” The sentence sounds universal, but the struggle underneath is still national power, corporate leverage, and who gets to set the terms of the machine age.
The piece said the NSA, FBI and CISA jointly accused six Chinese developers, including DeepSeek and Alibaba, of industrial-scale distillation of American frontier models earlier this month. It said Bessent said in July that Washington could sanction firms found stealing from American models. It said Trump and Xi are also expected to discuss whether Chinese firms can keep renting top Nvidia compute through Southeast Asian data centers, while the suspension of the Affiliates Rule, which extends export blacklists to affiliates of listed firms, expires with the Busan truce in November. The apparatus keeps expanding, and the people below it are expected to live with the consequences.
The analysis said the Trump administration has cleared roughly 10 Chinese companies to buy Nvidia’s H200 chips, though Chinese customs has kept nearly all the chips out. It said Beijing’s Foreign Ministry answered Amodei’s essay by warning that “fearmongering, confrontation and vicious competition” would disrupt global AI governance, while the state-run Global Times called the slowdown push a “Cold War playbook” for the sector. It said China’s Commerce Ministry threatened “resolute countermeasures” if Washington uses the distillation accusations to suppress Chinese firms, and that China could resurrect rare earth exports controls that were suspended under the Busan truce. Every side speaks the language of governance. Every side reaches for leverage.
The Channel They Say They Want
The piece said two of China’s most senior AI governance figures, Xue Lan and Zeng Yi, joined a Capitol Hill discussion convened by Senator Bernie Sanders this spring. It said Xue, who chairs China’s national expert committee on AI governance, called for “safe zones” where American and Chinese scientists can build safety standards and protocols together. It said Zeng, dean of the Beijing Institute of AI Safety and Governance, warned last week that AI’s catastrophic risks, including “the hidden danger of human extinction,” are mounting daily. It said both helped draft China’s national AI governance principles.
The analysis said Bessent proposed a standing China-U.S. AI dialogue in New York on September 20, with a way for each side to flag AI incidents serious enough to threaten national security. It said the AI consultations Trump and Xi discussed in Beijing in May were never formalized. It said U.S. Trade Representative Jamieson Greer said chip export controls were not part of the AI safety mechanism discussions. So the channel exists mostly as talk, while the real machinery of control keeps running elsewhere.
The piece said China-U.S. crisis management mechanisms already have a poor record, citing the February 2023 Chinese high-altitude balloon incident and Beijing’s later explanation that Washington had not created “the proper atmosphere” for dialogue. It said Amodei argued that restrictions “increase the leverage held by democracies and make an agreement more likely in the future,” and that he wrote that sharing information about misaligned models can have value even without formal agreements. That’s the old bargain: trust the institutions that keep the pressure on, and hope the pressure somehow produces safety.
The analysis said the Trump-Xi summit could produce a first substantive commitment that neither side will use AI to produce biological weapons. It said the two sides should establish a notification mechanism as a standalone safety channel, insulated from trade and technology-control disputes, and pair it with a technical track involving the Commerce Department’s Center for AI Standards and Innovation and experts like Xue and Zeng. It said Bessent testified before the House Financial Services Committee on September 15 and rejected the liability exemption the labs are seeking, arguing that safety is best guaranteed when developers are “liable for what they build and generate.” It said the independent evaluators that Anthropic and OpenAI have agreed to host do not require the United States to fall behind anyone.
The analysis concluded that the administration’s answer to calls to slow AI development is to force China to slow down instead, and said that may play well at home but will not make American AI safer and risks scuttling a China-U.S. channel on AI risk before it begins. That’s the whole game in one sentence: manage the optics, preserve the race, and call it protection.