Jensen Huang's Ouster Clears Anthropic's Biggest Threat Before $1.4 Trillion IPO
Huang's exit guts Nvidia's open-model ecosystem push days before Anthropic's IPO, which could value the AI lab at $1.4 trillion. Amodei's China stance is winning too.
By Grace Kim
4 min read

What's News
- Anthropic's IPO this week could value the company at $1.4 trillion.
- Nvidia accounted for just 39% of AI labs' future computing commitments, down from over 80% of AI chip revenue, per JPMorgan's Michael Cembalest on November 14.
- Nvidia developed open Nemotron models, backed Reflection AI, Thinking Machines Lab and Poolside, and moved to acquire Hugging Face.
- On Monday the US closed a loophole letting Chinese firms rent Nvidia GPUs from data centers outside China.
- Huang outlined his open-model vision at a Fireworks AI conference in San Francisco on November 3.
Jensen Huang's departure from Nvidia may be the best thing to happen to Anthropic just as the AI startup prepares for a stock market debut this week that could value it at $1.4 trillion.
Under Huang, Nvidia was investing heavily in cheaper, more open AI models that compete directly with Anthropic's offerings. The chip giant backed startups building these alternative models and helped make them widely available. That threatened Anthropic's core business: it spends heavily to develop closed models and charges customers to use them, often at higher prices than open alternatives.
"People really need to understand that the 'Cold War' going on here is that Nvidia was on a titanic collision course with the labs," the CEO of a leading AI startup told Business Insider, speaking anonymously to discuss sensitive topics.
The executive, who supports the open-source AI community and is sympathetic to Huang, described an ambitious Nvidia strategy: develop its own AI models, invest in rival startups, and build an ecosystem that could challenge not just Anthropic but OpenAI too.
That strategy was just beginning to come together. With Huang gone, the threat has diminished considerably.
What was Nvidia actually building?
The strategy centered on "open-weight" AI models — systems that, unlike Anthropic's Fable and Opus or OpenAI's top GPT models, can be downloaded, modified, and operated independently. Businesses like them because they get control without paying a leading lab on every use.
Nvidia pursued this aggressively:
- It developed its own family of open models, called Nemotron.
- It backed startups including Reflection AI, Thinking Machines Lab, and Poolside.
- It moved to acquire Hugging Face, a major platform for sharing open-weight AI models.
Replit CEO Amjad Masad called Huang's ouster "an attack on open models" at a Silicon Valley conference on Monday. "Jensen is the premier champion of open models," Masad said. "If it wasn't for Jensen, there wouldn't be a lot of funding going into US open model companies."
What did Huang say on stage?
On November 3, Huang evangelized the strategy at a Fireworks AI conference in San Francisco. He acknowledged the value of closed models from Anthropic and OpenAI, but argued businesses shouldn't have to rely on them.
"By taking an open frontier model, adapting it and providing it with your domain knowledge, your access, your specialized tools, and the proper evaluation, you should be able to create essentially a super intelligence," Huang told Fireworks CEO Lin Qiao on stage.
"And I think those times are here," he added.
The strategy made business sense for Nvidia: more open models running on its chips meant more customers. It also mattered as Anthropic and OpenAI sought alternatives to Nvidia's expensive hardware.
Why Anthropic was Nvidia's problem too
Anthropic has built what could be Nvidia's worst nightmare, as Business Insider reported in October. Its IPO filing discloses years of work on software that shifts AI work between chips from Nvidia, Google, Amazon, and AMD. That flexibility lets Anthropic pick the cheapest or most effective chip for each task, cutting its dependence on Nvidia and potentially weakening Nvidia's pricing power.
The shift is measurable. A November 14 analysis by Michael Cembalest of JPMorgan Asset Management estimated Nvidia accounts for just 39% of future computing commitments by leading AI labs — down from more than 80% of AI chip revenue in recent years.
Who won the China fight?
The two companies also clashed on China. Anthropic CEO Dario Amodei has advocated strict limits on advanced US chip sales to China, arguing they're essential to keeping America's technological lead. Huang fought those restrictions, warning that shutting China out would push Chinese firms to build competing technology.
Now Huang is gone, and Anthropic's worldview is prevailing. On Monday, the US government rolled out new restrictions closing a loophole that let Chinese firms use Nvidia GPUs by renting them from data centers outside China. That's a win for China hawks in the administration — and a huge victory for Amodei, arguably the biggest China hawk of them all.
There are caveats. Nvidia's interim CEO could continue Huang's strategies. Still, the immediate strategic implications are striking: days before Anthropic asks investors to bet on its future dominance of AI, the industry's most powerful figure pushing an alternative vision has been forced out.
Source: feeds.businessinsider.com
More from Grace Kim
Show full bio
Market editor covering people, launches and day-to-day practice in entrepreneurship, startups and venture funding at Business Bearings.
2,203 articles