The AI boom Enron parallels being drawn by analysts and investors have sharpened considerably with new data showing that off-balance-sheet obligations across major hyperscalers have ballooned to $3 trillion, all resting on AI revenue assumptions that remain unverified, according to Yahoo Finance. That context gives fresh weight to an analysis by Ram Bala, an associate professor of AI and analytics at Santa Clara University’s Leavey School of Business, who told Business Insider that today’s tech titans are replicating Enron’s financial playbook in three distinct ways, entirely legally, and with potentially far larger systemic consequences.
Three Enron Tactics, One AI Buildout
Enron, an energy supplier and trading powerhouse, collapsed into bankruptcy in 2001 after it was revealed to be a fraud. But Bala frames its collapse as a story of three specific financial manoeuvres: shifting debt off the balance sheet through special purpose vehicles, using mark-to-market accounting to book future sales projections as immediate revenue, and engineering circular transactions to make customer demand appear more independent than it was in reality.
‘The AI buildout is replicating versions of all three, entirely legally, and that is exactly why it deserves attention,’ Bala said.
On the first parallel, Bala said private credit is playing the same role that Enron’s special purpose entities once did, functioning as ‘the place where risk goes to become less visible.’ He pointed to private equity firms including KKR as participants in financing the AI buildout, and was direct about where the tail risk ultimately settles: ‘Nvidia gets paid upfront, the borrowers and their lenders carry the default risk, and because private credit is ultimately funded by ordinary savers through instruments like pension funds, the tail risk lands on households.’
On the second, Bala argued that the current buildout is being ‘financed against demand curves that are marked to model, not to market.’ He recalled Anthropic CEO Dario Amodei saying his company had invested less in compute than its rivals, given that a small error in demand projections could be the difference between success and bankruptcy. ‘When the smartest guys in this room are that candid about forecast fragility, leveraged borrowers underwriting to the optimistic case should give everyone pause,’ Bala said.
The AI Boom Enron Parallels Extend to Circular Financing
The third parallel is where the new data bites hardest. Bala described circular-financing arrangements such as Nvidia investing in OpenAI and OpenAI then using the cash to purchase Nvidia’s microchips as a form of vendor financing, common in capital-intensive industries, but risky ‘done too much with uncertain outcomes.’
The scale of these arrangements is larger than the report-level narrative suggests. According to RIA (Real Investment Advice), Nvidia has committed to investing $100 billion into OpenAI, with OpenAI’s CFO Sarah Friar acknowledging that ‘most of the money will go back to Nvidia.’ That is a textbook description of vendor financing, in which the supplier effectively lends the buyer the funds needed to purchase the supplier’s own product.
A separate transaction illustrates how layered these arrangements have become. Yahoo Finance reported that Nvidia invested $1.5 billion in SB Energy, funding the data centre that OpenAI will fill with Nvidia GPUs, with Nvidia effectively booking revenue twice on a single deal. Meanwhile, Axios has reported that Nvidia is reportedly considering a separate deal to finance OpenAI’s purchase of $350 billion in Nvidia chips. Taken together, these transactions represent a financing loop of a scale that dwarfs anything Enron constructed.
Bala noted that investors including Michael Burry have been among the loudest sceptics. Burry, who has 2 million followers on X, posted this week that ‘history is repeating’ and urged his audience to revisit the Enron story. In a separate post, Burry said: ‘This is so much bigger than one company, and orders of magnitude more dangerous to the economy and investors than Enron.’
Bala himself does not go that far. He believes the current boom is ‘not like bubbles of the past,’ and that long-term demand for AI will ultimately justify current financial strategies. The $3 trillion in off-balance-sheet obligations, however, means that the margin for error in those demand projections is extraordinarily thin, and that the consequences of being wrong would ripple well beyond the companies constructing the infrastructure.