Jensen Huang says Nvidia is pulling out of OpenAI and Anthropic, but his explanation raises more questions than it answers

At the Morgan Stanley Technology, Media and Telecommunications Conference in downtown San Francisco on Wednesday, Nvidia CEO Jensen Huang said his company’s recent investments in OpenAI and Anthropic would likely be the last in both, saying that once they go public, the opportunity to invest in “an important company like this closes.”

It can be that simple, yes. While companies sometimes pile into companies even on the eve of their public debuts in search of more upside, Nvidia is minting money by selling the chips that power both companies — it’s not as if it needs to boost its returns by pumping more money into either.

In fact, when asked for comment earlier today following Huang’s remarks, a TechCrunch spokesperson pointed to a transcript of Nvidia’s Q4 earnings call, where Huang said all of Nvidia’s investments are “directly and strategically focused on expanding and deepening the reach of our ecosystem” — something it has presumably already accomplished through its previous stakes in both companies.

However, some other dynamics may explain this decline. Industry observers have repeatedly pointed out that investing heavily in your largest clients creates circular and conflicting arrangements that can have negative downstream effects. When Nvidia first announced it would invest up to $100 billion in OpenAI last September, Michael Cusumano, a professor at MIT Sloan, described it to the Financial Times as “a bit of a wash,” noting that “Nvidia is investing $100 billion in OpenAI stock and OpenAI says it will buy $100 billion or more of Nvidia chips.”

This circular logic may help explain why Nvidia eventually scaled back this commitment. The investment it completed just last week, as part of a $110 billion round, reached $30 billion — far less than the $100 billion it once pledged. On Wednesday, Huang admitted as much, saying that investing the full amount was “probably not affordable.” Some have hypothesized that poor relations between the two companies could also be a factor, a suggestion Hwang called “nonsense.” Whatever the case, Nvidia’s relationship with Anthropic seemed fraught in itself.

Just two months after Nvidia announced a $10 billion investment in November along with a “deep technology partnership” with Anthropic, CEO Dario Amodei took the stage in Davos and, without mentioning Nvidia’s name directly, compared the actions of American chip companies selling high-performance AI processors to authorized Chinese customers to “selling nuclear weapons to North Korea.” (Those chip companies are Nvidia and AMD.)

It’s also worth noting what happened this week. Huang’s comments come just days after the Trump administration blacklisted Anthropic, barring federal agencies and military contractors from using its technology after the company refused to allow its models to be used in autonomous weapons or mass domestic surveillance.

Within hours of that announcement, OpenAI struck its own deal with the Pentagon — a move Anthropic called “false” and the public seemed to view it similarly. Within 24 hours, Cloud reached the top of the Apple App Store in the US, overtaking ChatGPT. (At the end of January, Anthropic was outside the top 100 companies, according to Sensor Tower data.)

That leaves Nvidia owning stakes in two companies that, at this particular moment, are headed in two very different directions — one newly allied with the Department of Defense, the other blacklisted by it.

It’s impossible to know if Huang saw any of this coming, given Nvidia’s network of partnerships. But the reason he announced Wednesday for possibly halting future investments — that the IPO window closes the door on these kinds of deals — is difficult to reconcile with the way late-stage private investment works.

What seems more likely, given all that has been revealed in recent days, is that this is a way out of a situation that has become really complicated, really fast.