MEXICO CITY, March 4, 2026, 15:11 (CST)
- Nvidia’s Jensen Huang signaled that the firm’s latest bets on OpenAI and Anthropic could be its last, with both AI players eyeing public listings.
- Huang said Nvidia has locked in a $30 billion stake, scrapping earlier talk of a possible $100 billion OpenAI investment.
- Washington is ramping up pressure over Anthropic’s military involvement, just as OpenAI looks into fresh defense-related applications.
Chief Executive Jensen Huang said Wednesday that NVIDIA Corporation’s recent investments in OpenAI and Anthropic may represent its final opportunities to secure major positions in the AI startups, with both companies eyeing an IPO later this year.
Nvidia’s been putting its cash to work, forging tighter links with firms fueling demand for its chips. When a company pivots from private fundraising to going public, that can block out last-minute strategic investors—potentially speeding up or complicating those connections.
At Morgan Stanley’s Technology, Media and Telecom conference, Huang said Nvidia has wrapped up its $30 billion investment, describing it as a rare opportunity to back “a consequential company like this.” A $100 billion bet on OpenAI isn’t likely, he added, especially with an IPO coming up. https://www.reuters.com/business/nvidia-wi…
Nvidia and OpenAI unveiled a $100 billion deal back in September last year, but now the company is calling the agreement more limited in scope. OpenAI has yet to provide a comment.
Huang signaled Nvidia’s $10 billion outlay in Anthropic is likely the company’s last, with the startup now considering an IPO. Anthropic, for its part, maintains it hasn’t made a final call on going public.
If Nvidia floats, that move would cast a sharper spotlight on the tangled AI ecosystem—especially since its largest clients often double as the major players behind today’s leading AI models. A few analysts have pointed out that hefty cross-ownership can start to look circular, particularly when profits from these firms ultimately end up funding more Nvidia chips.
Anthropic’s relationship with the U.S. government is already under scrutiny. An industry association counting Nvidia, Amazon and Apple among its members sent a letter to Defense Secretary Pete Hegseth, voicing “concern” over the idea of flagging Anthropic as a supply-chain risk. The group cautioned that such a move could “undermine the government’s access to the best-in-class products and services.” https://www.reuters.com/business/retail-co…
President Donald Trump last week rolled out a sweeping ban on Anthropic across federal agencies, giving them a six-month window to phase it out, according to Reuters. Hegseth also told Pentagon suppliers to strip Anthropic products from their supply chains.
OpenAI, in sharp contrast, is weighing a contract to roll out its technology on NATO’s “unclassified” systems, a person familiar with the matter said. That move comes just days after the ChatGPT developer landed a deal to put its tech into the Pentagon’s classified network. At a company meeting, CEO Sam Altman described the Pentagon agreement as “a complex, but right decision,” the Wall Street Journal reported, per Reuters. https://www.reuters.com/technology/openai-…
Nvidia faces more uncertainty with politics in the mix. If regulators crack down on a key customer’s applications, or if a messy debate erupts over guardrails, chip demand could take a hit. That would feed right into how investors weigh Nvidia’s riskier plays.
There’s a more basic risk here too: IPO timelines can get pushed back, valuations might drop, or regulators could step in. What seems like a savvy stake during a hot market can quickly become a drag if the funding environment shifts.
Huang’s remarks point to Nvidia carving out its boundaries—supplying the “picks and shovels,” but stepping back from deeper involvement with clients headed for the public markets spotlight.
