In brief
- Nvidia signaled the end of its major investments in OpenAI and Anthropic.
- The OpenAI-Anthropic feud is escalating fast after a Pentagon deal and federal backlash.
- With this decision, Nvidia avoids choosing sides as AI labs clash with each other and Washington.
At the Morgan Stanley Technology, Media, and Telecom conference in San Francisco on Wednesday, Jensen Huang said Nvidia is likely done investing in AI giants OpenAI and Anthropic.
The $30 billion investment it just finalized into OpenAI was a steep drop from the $100 billion deal announced last September, and will likely be the last check. Same for the $10 billion it put into Anthropic in November.
The official reason sounds clean: Both companies appear to be heading toward IPOs later this year, and private deals of this nature close once that window opens.
“This might be the last time we’ll have the opportunity to invest in a consequential company like this,” Huang told the audience.
Granted, late-stage investors write checks right up to the IPO bell all the time, and the original $100 billion OpenAI commitment didn’t shrink to $30 billion because of some procedural IPO rule. Something else changed.
Nvidia now holds stakes in two companies in the middle of an all-out war with each other—and with Washington. Trump ordered all federal agencies to stop using Anthropic’s technology after the company refused to let Claude be deployed for autonomous weapons or mass domestic surveillance. Within hours, OpenAI announced its own Pentagon deal—a move Anthropic publicly called “mendacious.”
The public sided with Anthropic. Within 24 hours of the back-to-back announcements, Claude shot to the top of the free app rankings on Apple’s U.S. App Store, overtaking ChatGPT. At the end of January, it was outside the top 100. An Anthropic spokesperson told Decrypt that it saw record signups in the days following the Pentagon’s move.
Meanwhile, the QuitGPT movement had claimed an estimated 2.5 million users taking action against OpenAI—canceling subscriptions or spreading the boycott—by the time the dust started settling.
Nvidia’s relationship with Anthropic was already strained before all this. Two months after putting $10 billion into the company, Dario Amodei stood at Davos and compared U.S. chip companies selling high-performance processors to approved Chinese customers to selling nuclear weapons to North Korea. He didn’t name Nvidia, but didn’t need to either.
There’s also the structural awkwardness of the whole circular economy behind Nvidia’s massive investments in AI startups. Nvidia invests in OpenAI, but OpenAI spends it on Nvidia chips. The circularity drew bubble comparisons.
What Nvidia is actually doing is getting out of the business of picking sides. It sells GPUs to OpenAI, Anthropic, xAI, Google, and everyone else racing for the frontier. The arms dealer doesn’t get to have a favorite army.
Getting caught holding equity on both sides of a Pentagon standoff, in which one investee is increasingly hated and the other gets designated a national security supply-chain risk, is exactly the kind of mess that makes customers nervous.
The IPO story is a convenient door. Huang walked through it.
