The Omniscient Bulletin · November 16, 2026

Washington presses Nvidia over its next CEO as allies sign an AI supply-chain declaration


Nvidia's leadership crisis has turned into a policy fight: Treasury Secretary Scott Bessent wants an outsider as chief executive and answers on export compliance within 10 days, while the board says it has not yet chosen a search firm. Separately, a Reuters review found more than $3 billion in Chinese projects built around restricted Nvidia chips. Elsewhere, the US and nine partners signed a supply-chain declaration, OpenAI gave its agents a wallet, and Alibaba made Qwen4-Max generally available.

Policy

Treasury Secretary Scott Bessent said Nvidia should look outside the company for its next CEO and wants answers on its export-control compliance within 10 days

Treasury Secretary Scott Bessent said Sunday on Fox News that Nvidia's board should go outside the company as it searches for a new chief executive, arguing that it needs "a dramatic shake-up" after Jensen Huang stepped down last week. He also said the board and its outside counsel must answer questions within 10 days about Nvidia's compliance with export restrictions. Huang resigned Thursday after reports of affairs, one of them with a suspected Chinese spy.

Fox Business →


Industry

Nvidia's board said it is still choosing a search firm for its next CEO and has not formally engaged any candidates, as the FT reported it approached Arm CEO Rene Haas

Nvidia's board said Friday it is still choosing a search firm to find a permanent chief executive and has not formally engaged any potential candidates, calling talk of individuals premature. The Financial Times reported the same day that the board had approached Arm CEO Rene Haas, who spent seven years at Nvidia before joining Arm in 2013. Interim CEO Colette Kress held her first meeting with top executives on Thursday. The stock dropped more than 2% Friday, pushing Nvidia's market value under $4 trillion.

CNBC →


Policy

The US and nine partners, including Japan, India and the UK, signed a declaration on protecting the technologies and infrastructure that AI supply chains depend on

On Sunday, the United States and nine partners in its Pax Silica coalition signed a declaration setting out a shared view of which sensitive technologies and critical infrastructure underlie Super Intelligence (SI) supply chains (for some signers, the text notes, SI means AI) and agreeing that protecting them is a shared task. The other signers are Australia, India, Israel, Japan, South Korea, Qatar, Singapore, the UAE and the UK. They proposed five working groups, one of them on SI security and resilience.

US State Department →


Agents

OpenAI launched ChatGPT Wallet, which lets its agents pay for purchases without seeing users' card details, starting with ChatGPT Pro subscribers in the US

OpenAI introduced ChatGPT Wallet, built with American Express, Chase, Mastercard and Visa, so agents can pay on a user's behalf without access to the actual card details. For every purchase, the user checks the items, total and card, then approves it with Face ID, a fingerprint or a similar check, and the payment network issues a substitute token. It starts in dots, OpenAI's agents, for US ChatGPT Pro subscribers. OpenAI expects users eventually to delegate purchases within spending limits or other rules they set.

OpenAI →


Compute

A Reuters review found more than $3 billion in Chinese AI computing projects built around Nvidia processors that US export controls restrict

Reuters reviewed over 150 Chinese procurement notices, contracts, investment plans and disclosures and identified projects and equipment deals worth over 23 billion yuan ($3.2 billion) built around Nvidia's restricted H100, H200 and H800 chips. The tally mixes planned, contracted and finished work and includes buildings and power systems, not just chips. Nvidia has said export rules effectively shut it out of China's data center market. Reuters did not link the projects to Megaspeed, the Singapore-based Nvidia cloud partner, or to the allegations involving Huang.

Reuters via WTVB →


Frontier

Alibaba made its flagship Qwen4-Max model generally available, a 2.4 trillion-parameter system with a context window of up to one million tokens

Alibaba on Monday made Qwen4-Max, its flagship model, generally available through QwenCloud after a limited preview that began November 11. It is a mixture-of-experts model with 2.4 trillion parameters that accepts images directly and handles up to one million tokens of context. Alibaba says it uses fewer tokens than Qwen3.8-Max, which together with lower API prices cut the average task cost on its own Work360 benchmark by about 58%. The company also plans to open-source smaller models in the Qwen4 series.

TechNode →


Industry

OpenAI disbanded its stand-alone robotics team and moved its staff into its main research and applied groups, Fortune reports

OpenAI has broken up its stand-alone robotics team, sending researchers to the main research group under Chief Research Officer Mark Chen and applied and hardware staff to the Applied organization under President Greg Brockman, Fortune reported. A spokesperson said robotics remains a core part of its long-term vision. Researchers have seen GPT-6 Astra do surprisingly well when connected to robots, Fortune said. OpenAI has said it is building its own humanoid robot, targeting a first public demonstration in 2027.

Fortune →

Get every weekday issue in your inbox.

All issues →