Did Circular Deals Prop Up the Artificial Intelligence Boom?

2026-11-14 companies

Santa Clara, Saturday, 14 November 2026.
Nvidia faces allegations of artificially prolonging processor scarcity by committing hundreds of billions to venture-backed startups that pledged to buy surplus chips major cloud providers no longer wanted.

A Leadership Transition Amidst Scrutiny

Nvidia (NASDAQ: NVDA) finds itself at a critical inflection point as former CEO Jensen Huang exited abruptly this month, following reports of extramarital affairs, one involving a suspected Chinese spy [5]. His resignation, effective November 11, 2026, immediately elevated CFO Colette Kress to interim CEO, while Harvey C. Jones was appointed the company’s first-ever Chair of the Board [4]. The transition follows a period of intense scrutiny and alleged strategic missteps, including a reported internal rift over circular financing arrangements. Earlier this year, Kress had pushed to reduce $250 billion in commitments to OpenAI and other AI startups—only to be rebuffed by Huang, who insisted on doubling down on such deals to sustain GPU demand [2]. This internal conflict now casts a shadow over Nvidia’s reported strategy of using venture investments to artificially sustain demand for its proprietary hardware.

Circular Financing: A Two-Edged Sword

At the heart of the controversy is a web of circular financing deals, totaling approximately $530 billion, designed to prop up demand for Nvidia’s GPUs by funding AI startups and cloud providers that lacked the capital to purchase the chips outright [2]. Documents reviewed by The Wall Street Journal describe a high-level meeting in August 2026, during which Huang berated Kress for questioning the sustainability of these arrangements. Huang argued that such guarantees were necessary to prevent Nvidia from being left holding an “avalanche” of unsold chips from TSMC, as major hyperscalers began shifting to custom silicon [1][2]. The irony, as reported, is that many of these funded startups—such as CoreWeave, where Nvidia holds a $2.7 billion stake—were unlikely to generate sufficient revenue to cover their commitments [2][3]. As one industry observer noted, these deals were less about market economics and more about maintaining the illusion of scarcity in an overheated AI market [alert! ‘Huang’s motivations remain unconfirmed outside internal accounts’].

The Data Center Dilemma: Supply vs. Demand

Nvidia’s strategy hinged on the assumption that AI startups and smaller cloud providers would eventually scale into major customers. However, internal documents reveal growing unease within the finance team, with Kress warning that Nvidia’s receivables had ballooned to $63 billion, while average payment terms stretched from 45 to 60 days—a signal that customers were struggling to meet their obligations [2]. Meanwhile, major cloud providers like Google and Microsoft were rapidly developing in-house AI chips, reducing their reliance on Nvidia’s proprietary GPUs [GPT]. The result, according to a WSJ investigation, was a classic case of channel stuffing, where Nvidia used venture funding as a crutch to absorb excess silicon that otherwise would have flooded the market [1]. Critics argue this not only distorted pricing but also delayed the inevitable correction in AI hardware costs.

Regulatory and Market Repercussions

The allegations have reignited antitrust concerns, with analysts questioning whether Nvidia’s practices violated fair competition principles by artificially constraining supply to inflate prices [alert! ‘no formal investigations have been publicly confirmed’] The timing of Huang’s departure—and the abrupt shift in leadership—raises questions about accountability. Nvidia’s upcoming earnings report on November 17, 2026, will serve as a crucial test for Kress’s interim leadership, particularly as investors weigh the long-term viability of the company’s circular financing model [4][6]. Competitors like AMD and Broadcom, which have focused on ecosystem partnerships rather than capital-intensive vertical integration, appear better positioned to weather a potential downturn in AI hardware demand [2][GPT]. As the dust settles, one thing is clear: the AI boom’s foundation may have been shakier than previously believed.

Sources

Artificial intelligence Semiconductor industry