Sacks, on All-In, ties Huang’s Nvidia exit to a fight over who controls AI models and chips

Documented: On November 12, 2026, Nvidia’s board said Jensen Huang had resigned as president, chief executive, and a director, effective immediately. Colette Kress, the company’s CFO since 2013, was named interim CEO while the board searches for a permanent successor, and Harvey C. Jones was named the company’s first board chair. Reuters and TechCrunch reported the resignation came days after Huang admitted an extramarital affair and after reporting that one relationship involved a Chinese businesswoman, Alice Huang (also referred to as Huang Le), whom U.S. authorities had suspected of ties to Chinese intelligence. Nvidia shares fell sharply over the week; Reuters described the exit as a shock to the AI boom Huang had helped finance and lobby for, including his push to keep some chip access to China.

What is not established: that Alice Huang was a spy, or that the board removed Huang because of open-source politics. Reuters said it could not verify the espionage allegations. Beijing has denied she is a spy. CNN reported that U.S. intelligence had not proven she was acting for China’s Ministry of State Security, and that officials disagreed about the strength of the evidence; Attorney General Todd Blanche said the Justice Department had reached no conclusion about who she is. A WIRED report, citing a person familiar with the matter, said the alleged relationship with a direct report and Huang’s dealings with Singapore-based Megaspeed likely mattered more to the board than the espionage claims.

Claimed, on the November 14 All-In episode “Jensen Ousted as Nvidia CEO”: David Sacks, the former White House AI adviser who co-chairs the President’s Council of Advisors on Science and Technology, argued that “safetyist” groups aligned with and funded by Anthropic were celebrating Huang’s fall, and that the real agenda was extreme centralization—government control of models and chips, with surveillance and back doors. Hosts framed Huang as the main corporate backer of open-weight AI and as someone who had opposed kill switches and location tracking on GPUs. A Chorus summary of the episode records Sacks saying safety groups “don’t really care about safety” and that open-source models are what actually limits closed-model centralization. Those remarks are commentary by a political and industry figure, not a finding by Nvidia, Congress, or an investigator.

The institutional fracture underneath the podcast theory is real even if the causal story is not. Documented policy fights already split the stack: export controls and location-tracking proposals on advanced chips; closed frontier labs versus open-weight models that have narrowed the capability gap at lower cost; and a safety agenda that, in Sacks’s telling and in earlier All-In discussions of Anthropic, would treat access to powerful models as something the state should ration. Anthropic has been moving toward a public listing; whether that IPO, or any political spending on safeguards and export controls, was timed to Huang’s resignation is speculation. No public record reviewed here shows Anthropic directed the leak of a classified memo, engineered the affair disclosures, or dictated the board vote.

Bracketed cleanly: Huang is out, by the company’s own statement, after a personal and national-security scandal whose spy element remains unproven in public. The market and export-policy consequences of losing the founder who personified Nvidia’s China and open-ecosystem stance are a documented risk analysts and Sacks both flag. The claim that “safetyism” was the weapon that took him out—so that Anthropic’s allies could seat the government over models and chips—is a theory offered on a podcast. It is interesting because it names a live power struggle over compute. It is not evidence that struggle caused the resignation.