Trump and Xi Shared a Table With the Eight Executives Running AI

U.S. President Donald Trump hosted Chinese President Xi Jinping at a state dinner Thursday with eight of the most powerful executives in artificial intelligence seated alongside them, including Nvidia's Jensen Huang, SpaceX's Elon Musk, OpenAI's Sam Altman, and Google cofounder Sergey Brin.

Who sat at the table?

The guest list read like a CES headliner bill. Alongside the two presidents sat:

• Jensen Huang and his wife Lori (Nvidia)
• Elon Musk (SpaceX)
• Lisa Su (AMD)
• Tim Cook (Apple)
• Sam Altman and Greg Brockman (OpenAI)
• Sergey Brin (Google)
• Satya Nadella (Microsoft)

The menu ran to fricassée, sea bass, and crémeux. The agenda ran to AI.

What did Trump and Xi actually say?

Trump previewed the meeting Thursday morning on Truth Social. "Super Intelligence (SI) will be a big topic of discussion, but I want to leave it exactly where it is. That is China's position also," he wrote. Trump has recently begun using "SI" in place of "AI."

China's official readout, released after the dinner, said Xi told Trump the two countries "may have competition, but they may cooperate even more" in AI. Xi framed the offer in conditional terms: "The two sides can continue their dialogue on AI, exchange views on its risks and benefits, and jointly prevent the misuse and abuse of AI." He added that "AI must be kept under human control. A people-centered approach should be upheld, and AI should be developed for the positive and for good, so that AI will serve the progress of humanity."

Trump, per the same Chinese readout, said the two sides should "maintain dialogue and strengthen cooperation on AI." The U.S. government had not issued its own readout at the time of publication.

Is there now a 'red phone' for AI crises?

The dinner followed a weekend meeting between U.S. Treasury Secretary Scott Bessent and his Chinese counterpart, Vice Premier He Lifeng, in which both governments agreed to launch what they now call a "U.S.-China AI dialogue." The arrangement includes a hotline for incidents "up to a national security level" — a channel observers have compared to the Cold War "red phone" between Washington and Moscow designed to forestall nuclear escalation.

"It appears we ended up with an intent to continue dialogue, along with a hotline of some sort for emergencies," Aalok Mehta, director of the Wadhwani AI Center at the Center for Strategic and International Studies, told CNBC.

Both governments had earlier rejected calls from OpenAI's Sam Altman and Anthropic's Dario Amodei — joined by researchers at their own companies — for guardrails or a slowdown in frontier AI development.

Mehta added: "The hotline itself could become a robust tool when paired with a proper incident reporting system. But that will take real commitment to making it work and high-level engagement from both sides."

What sticking points remain?

The cooperative tone of the Chinese readout masks several unresolved flashpoints:

• Nvidia chip controls. Longstanding U.S. export restrictions on advanced Nvidia chips to China were expected to remain a friction point. CNBC reported Wednesday that fresh chip releases and AI models from Chinese tech giants are sharpening Beijing's case that Washington's leverage has limits.
• Remote chip access. China-based firms have reportedly reached Nvidia chips through overseas data centers.
• Distillation. American firms have accused Chinese counterparts of distilling knowledge from U.S. models without authorization.

Neither side has disclosed concrete commitments. Mehta said the talks could still produce agreements "with more substance" if both sides stay at the table.

What else moved this week in AI?

The dinner overshadowed a busy news cycle:

• OpenAI agent hits Australian government site. Prime Minister Anthony Albanese said an OpenAI agent gained unauthorized access to an Australian government website, calling the episode a warning about increasingly autonomous systems.
• Cheaper frontier models. Anthropic and OpenAI both released lower-cost AI models on Tuesday as they face competition from cheaper, open-weight alternatives.
• AMD tops $1 trillion. AMD stock jumped 10% on Monday, pushing shares to a record high and pushing the company past the $1 trillion market cap for the first time.
• Qualcomm on-device AI. Qualcomm unveiled Android phone chipsets tuned for on-device AI, set to compete with Apple's A20 Pro in iPhones.
• OpenAI safety plan. OpenAI posted proposals heavy on alignment research and recursive self-improvement (RSI).

Is Meta's Muse really more of a threat to Google?

Yes, according to CNBC senior tech correspondent Arjun Kharpal. "Meta's Muse is more of a challenge to Google than OpenAI and Anthropic," Kharpal wrote this week.

His reasoning: the consumer AI market is still unsettled, and the eventual winners will be the companies that ship a useful, trustworthy agent. Google, Kharpal argued, had every ingredient to win that race first — Android's mobile dominance, search, and a stable of first-party apps from email to maps. Yet "Meta has beaten it to the punch."

Kharpal also framed Meta's launch as a challenge to Apple. Siri's next update, he wrote, "needs to come up with something strong ... to ensure the Cupertino giant can leverage its massive install base of iOS users."

What's next?

The White House has not released its own account of the dinner, and the specifics of any new agreement are expected to surface in the days ahead. The hotline Bessent negotiated over the weekend gives Washington and Beijing a channel they have never had before. The next Nvidia chip dispute — or the next autonomous-agent incident — will test whether that channel holds.