Nvidia has already committed $40 billion to equity AI deals this year - and it's reshaping the ecosystem

Nvidia has poured more than $40 billion into equity stakes in AI companies since the start of 2026 - a pace, in just four months, 75% above the $23 billion it deployed in all of 2025. Most of those investments are flowing to companies that are simultaneously customers of its GPUs, in what Nvidia describes as "strategic balancing" capital contributions.
Nvidia's largest publicly disclosed moves so far this year include a $6.5 billion investment in xAI (March 2026), an additional $4 billion in CoreWeave (February 2026, bringing the cumulative position to $9 billion), $2.8 billion in Mistral AI (April 2026), $1.9 billion in Lambda Labs (February 2026) and $700 million in Argentine cloud provider Siglo (March 2026). Of those investments, roughly $32 billion is going to companies expected to be major buyers of Nvidia's GPUs over the next three years.
The structure raises an accounting concern that critics call "compromised circular reserves." On the income side, Nvidia books GPU orders from its investees as revenue; on the balance sheet, the investments themselves are recorded as capital assets. Investors are openly asking whether the structure inflates Nvidia's true operational performance.
A Financial Times analysis last month estimated that approximately 18% of Nvidia's revenue growth in 2025-26 came from these kinds of cross-deals. Nvidia CFO Colette Kress rejected the characterisation on the most recent investor call: "Our investments are responsive to market dynamics. Our customers' preference for our GPUs is a fact independent of our equity-investment position."
Former head of the US antitrust division Jonathan Kanter told the FT that "Nvidia's investment strategy is the kind of model that should be watched carefully on antitrust grounds. A single supplier directing the capital flow into its customer ecosystem can in turn restrict customers' supplier choices." The Department of Justice opened a formal investigation in 2024 into Nvidia's control of the AI ecosystem; the investigation is still openly ongoing.
The combined post-money valuation of companies in which Nvidia has invested in 2026 stands at over $480 billion. That is equal to roughly 31% of the global AI ecosystem - markedly ahead of Microsoft (22%), Google (14%) and Amazon (11%). The equation between capital concentration and market share gives important clues to the sector's future competitive structure.
Investments at this scale are leaving an imprint on Nvidia's balance sheet. The company's first-quarter 2026 balance sheet shows capital investments at $89 billion, four times last year's first quarter. Its cash and cash-equivalent holdings have dropped within the year from $145 billion to $92 billion. AAII investor advocate Charles Rotblut said: "These numbers indicate Nvidia's true operational risk profile is at its highest level in three years."
Other major tech firms are reacting differently to Nvidia's structure. AMD CEO Lisa Su said at Computex last month: "Investing equity in our customers is not our business model. Supplying them with chips that perform consistently well is our business model." AMD's AI-chip market share rose to 12% in the first quarter of 2026 (from 8% a year earlier), pushing investors to look for a counterweight to Nvidia's dominance.
A further question is the sustainability of Nvidia's pace. A $40 billion total over four months exceeds Silicon Valley's six largest VC ecosystems combined. If Nvidia maintains this pace, its investments by year end could reach $120 billion - roughly two-thirds of all the capital entering AI start-ups globally in 2025.
In Nvidia's May 2026 investor presentation, CEO Jensen Huang stated the rationale plainly: "Scaling AI infrastructure is achieved not just by producing chips but by populating the entire ecosystem with strong partners. These investments are part of our strategy to spread the power of our GPUs into the service of the world as fast as possible." Critics counter that this "populate-the-ecosystem" strategy carries the risk of making Nvidia AI's central capital provider. The answer once again depends on whether the FTC or the European Commission decide the matter requires a seat at the table.