Jensen Huang says Nvidia is moving away from OpenAI and Anthropic, but his explanation raises more questions than it answers.

At the Morgan Stanley Technology, Media and Telecom conference in San Francisco on Wednesday, Nvidia CEO Jensen Huang said that his company’s recent investments in OpenAI and Anthropic should be the last in both companies, saying that once they go public, the opportunity to invest in “the next company like this” closes.

It could be that simple, yes. Even companies are sometimes grouped into companies until approx the first day of their public exhibition in search of more, Nvidia is making money to sell chips that power both companies – it’s not like it needs to pay back its returns by pouring more money into one.

Instead, he was asked to comment today following Mr. Huang’s statement, a spokesperson pointed TechCrunch to a note from the fourth quarter of Nvidia’s earnings, where Huang said that all of Nvidia’s businesses are “focused, technically on the expansion and depth of our ecosystem” – which is likely to have already achieved its companies.

However, several other variables could also explain the pullback. Industry executives have repeatedly said that investing heavily in your biggest customers creates roundabout, counterproductive strategies that can have negative consequences. When Nvidia first announced that it would invest $100 billion in OpenAI last September, MIT Sloan professor Michael Cusumano explained to the Financial Times that “like bathing,” noting that “Nvidia is putting $100 billion in OpenAI stock and OpenAI is saying it will buy $100 billion or more of Nvidia’s chips.”

This circular reasoning may explain why Nvidia pulled back on the commitment. The money that ended last week, as part of the round of $ 110 billion, came $30 billion – a shortfall of $100 billion previously pledged. On Wednesday, Huang conceded more, saying the investment “may not be in the cards.” Some have suggested that bad blood between the two companies may also be a reason, a suggestion Huang has called “nonsense.” Either way, Nvidia’s relationship with Anthropic appears to be strained on its own.

It’s been two months since Nvidia announced it $10 billion investment in November along with a “deep technical cooperation” with Anthropic, CEO Dario Amodei took part in Davos and, without naming Nvidia directly, he compared the US chip companies selling advanced AI processors to convince Chinese customers that “selling nuclear weapons to North Korea.” (Those chip companies are Nvidia and AMD.)

It is also worth knowing what happened this week. Huang’s comments come just days after Trump’s inauguration not allowed Anthropic, barring government agencies and military contractors from using its technology after the company refused to allow its models to be used as autonomous devices or for mass surveillance at home.

Just hours after the announcement, OpenAI entered into a partnership with the Pentagon – a move Anthropic called “mendacious” and it seems that people felt the same way shot up of Apple’s US App Store, beyond ChatGPT. (At the end of January, Anthropic was outside the top 100, according to Sensor Tower data.)

Where it leaves off Nvidia is working in two companies that, at the moment, are pulling in very different directions – one that has just cooperated with the Department of Defense, and the other that has been registered with it.

Whether Huang saw this coming, given Nvidia’s contract page, is impossible to know. But Wednesday’s reason for wanting to pull the plug on future investments — that the IPO window will close the door on such a deal — is that it’s hard to know how late-stage private equity investment works.

What seems likely, given all that has happened in recent days, is that this is getting out of a very difficult situation, very quickly.