Nvidia’s stock slid more than 5% on Monday, closing at €172.10, as investors grappled with a paradox: the chipmaker is deepening its commitment to artificial intelligence through a high-profile startup investment while simultaneously exploring a role as guarantor for its own customers’ infrastructure projects. The dual narrative — one of strategic partnership, the other of financial exposure — left the market parsing risk from opportunity.
A Startup Bet Without Revenue
The company announced a long-term strategic partnership with Safe Superintelligence Inc. (SSI), the research lab founded by former OpenAI chief scientist Ilya Sutskever. Nvidia is making a “significant” investment in SSI, though neither party disclosed the exact figure. In exchange, the lab gains access to Nvidia’s upcoming Vera Rubin computing platform. Co-founder Daniel Levy projects that SSI’s computational capacity will increase tenfold within twelve months.
The deal is notable because SSI has neither a finished product nor any revenue. Yet through this arrangement, it effectively leapfrogs into the same league as major cloud providers when it comes to securing high-end chips. The partnership underscores Nvidia’s willingness to place long-term bets on frontier AI research, even when the commercial payoff remains uncertain.
Alongside the SSI tie-up, Nvidia co-founded the Open Secure AI Alliance alongside Microsoft, IBM, Cisco, and SAP. The initiative aims to develop open-source tools for auditing AI agents and protecting open models, with a focus on standardized formats for model weights and digitally signed software patches. The alliance follows an open letter from the industry urging regulators to target concrete harms rather than impose blanket restrictions on open AI infrastructure.
The $250 Billion Question
The stock’s decline, however, was driven by a different story entirely. Reports emerged that Nvidia is in early-stage negotiations to provide a $250 billion financing guarantee for a massive data center project in Ohio — a 10-gigawatt facility to be built by SoftBank and leased by OpenAI. A separate chip financing arrangement worth $350 billion is also under discussion, meaning Nvidia could ultimately be on the hook for up to $600 billion in credit exposure.
The logic behind the move is straightforward: traditional credit markets are hesitant to finance OpenAI’s $500 billion infrastructure ambitions. OpenAI lacks an investment-grade rating and is projected to post a $14 billion loss in 2026. By stepping in as a guarantor, Nvidia ensures that its own hardware continues to be purchased at scale, even when customers struggle to secure conventional financing.
Should investors sell immediately? Or is it worth buying Nvidia?
But the arrangement raises uncomfortable questions. Critics, including noted investor Michael Burry — who is reported to have expanded his short position against Nvidia — warn of a dangerous feedback loop. Nvidia would effectively be guaranteeing the debt that customers use to buy Nvidia’s own chips. The strategy works only if AI software profitability materializes at the scale currently anticipated. OpenAI is expected to accumulate cumulative losses of $115 billion by 2029, while ChatGPT’s share of global web traffic has already slipped from 86.7% in January 2025 to 64.5% in January 2026. That divergence suggests end-user demand may be growing more slowly than the financed infrastructure implies.
Bullish vs. Bearish: Two Views of the Same Risk
Proponents see Nvidia’s balance-sheet deployment as a logical evolution. The company is effectively securing its own revenue stream by ensuring that capital-intensive projects proceed despite tight credit conditions. A recent example: Nvidia invested $1 billion for a 4.5% stake in South Korea’s Naver to support its “AI Factory” expansion. Global AI buildout is estimated to consume $7.5 trillion over the next five years, and Nvidia’s willingness to act as a capital partner locks in long-term demand. Bank of America maintains a buy rating with a price target of €266.21, implying roughly 54% upside from current levels.
Skeptics counter that the strategy introduces systemic concentration risk. Nvidia is moving from selling chips to underwriting the projects that buy them — a shift that transforms its risk profile. The stock now sits 14.71% below its 52-week high of €202.50 from May, and the market is beginning to price in these contingent liabilities.
Technical Picture and What Comes Next
Nvidia’s shares have fallen 15.01% from that May peak and now trade below their 50-day moving average of €180.45. The relative strength index sits at 42.7, approaching neutral territory after a sustained pullback. The 200-day moving average of €166.18 remains intact, offering a technical floor roughly 4% below the current price.
The immediate catalyst for the stock’s direction lies in earnings reports from Nvidia’s largest customers. Microsoft and Alphabet are both due to report this week, and investors will scrutinize whether their massive capital expenditures are translating into concrete orders for Nvidia’s Blackwell and Vera Rubin systems. If cloud giants fail to show accelerating demand, the pressure on Nvidia’s share price is likely to persist.
Further out, the key inflection point will be whether the $250 billion guarantee talks with OpenAI are confirmed or collapse. That decision will determine whether Nvidia formally enters the world of large-scale credit provision. A more distant milestone is the first 800-megawatt phase of the Ohio project, slated for 2028, which will test whether this capital-intensive strategy actually delivers returns. For now, the market is watching, weighing guaranteed chip sales against the weight of new contingent liabilities.
Ad
Nvidia Stock: Buy or Sell?! New Nvidia Analysis from July 27 delivers the answer:
The latest Nvidia figures speak for themselves: Urgent action needed for Nvidia investors. Is it worth buying or should you sell? Find out what to do now in the current free analysis from July 27.
Nvidia: Buy or sell? Read more here...
