OpenAI, the company that ignited the artificial intelligence (AI) boom with ChatGPT in 2022, faces a financial predicament that dwarves even the harshest level of criticism against it.
The AI pioneer needs to secure at least $207 billion in additional funding by 2030 simply to continue operating at a loss, according to recent analysis from HSBC’s software and services team.
The numbers paint a stark picture of unsustainable growth. OpenAI recently committed to a $250 billion rental agreement with Microsoft Corp. and a $38 billion contract with Amazon.com Inc. for data center capacity. HSBC projects these commitments will translate to annual rental costs of $620 billion, potentially ballooning to $1.4 trillion by 2033 — exceeding Saudi Arabia’s GDP.
The spending spree stands in sharp contrast to the company’s revenue trajectory. CEO Sam Altman said OpenAI’s annualized revenue is approaching $20 billion in 2025 – less than 10% of the $288 billion total cost of its Microsoft and Amazon data center agreements alone. The compute costs projected for 2033 would be roughly 70 times current revenues, before accounting for staffing, research and development, energy, water, or property expenses.
The business model presents a fundamental challenge: Only about 5% of ChatGPT’s user base currently pays for subscriptions, yet these paying subscribers account for approximately 70% of annual recurring revenue. HSBC estimates OpenAI would need to reach three billion users by 2030 and convert 10% to paid subscriptions to approach sustainability — ambitious targets given that user growth appears to be plateauing around 800 million weekly active users.
Competition adds another layer of pressure. Google’s recent Gemini 3 launch has intensified the battle for market share, with HSBC predicting OpenAI’s consumer dominance will decline substantially by decade’s end. Meanwhile, dozens of competing platforms are vying for the same limited data center capacity, potentially driving costs even higher.
The company’s partners are shouldering enormous debt to support OpenAI’s expansion. An analysis by the Financial Times reveals suppliers of data centers, chips, and computing power to OpenAI have accumulated approximately $96 billion in debt. This includes $30 billion borrowed by SoftBank Group, Oracle Corp., and CoreWeave, and $28 billion in loans taken by Blue Owl Capital and Crusoe, with another $38 billion in potential financing under discussion.
CoreWeave’s situation exemplifies the precarious nature of OpenAI’s debt-fueled growth. The company reported $14 billion in current and non-current debt, plus $39.1 billion in future lease agreements, while projecting just $5 billion in revenue for this year. The five major hyperscalers — Amazon, Google, Meta Platforms Inc., Microsoft, and Oracle — have collectively taken on $121 billion in new debt this year to fund AI operations, more than quadruple their average debt issuance over the previous five years.
Credit markets are beginning to respond to these mounting risks. Oracle’s five-year credit default swap spreads have widened by approximately 60 basis points since late September, while CoreWeave’s have increased by roughly 280 basis points, signals that investors are growing concerned about default risks.
Most ominously, copyright litigation threatens the entire foundation of OpenAI’s business. Following Anthropic’s $1.5 billion settlement with authors over scraped copyrighted content — essentially $3,000 per document — approximately 50 lawsuits worldwide target AI companies for alleged copyright theft. OpenAI faces multiple cases related to unauthorized training data, while its planned expansions into music and video services could trigger additional legal challenges given that users can already generate output featuring copyrighted characters and Hollywood actors.
To ease financial pressures, OpenAI is reportedly pivoting toward advertising revenue, raising privacy concerns. The company would monetize the intimate details users share with ChatGPT — family information, health anxieties, career struggles, and other deeply personal data. Critics have characterized this as a betrayal of user trust.
Nonetheless, OpenAI has its defenders on Wall Street.
Big Tech companies will continue to dominate market performance in 2026 as AI and real-world applications take center stage, according to Dan Ives, an analyst at Wedbush Securities. He forecasts tech stocks will rise another 20% next year as the AI revolution enters its next phase, driven by what he calls “2nd/3rd/4th AI Revolution derivatives trades.”
The analyst rejects characterizations of an AI bubble, pointing to five key factors that suggest the technology’s growth trajectory remains in its early stages. The consumer AI revolution has yet to materialize, he notes, while autonomous vehicle technology is just beginning to emerge and robotics applications remain largely confined to laboratories though poised for mainstream adoption within years.
Ives also emphasizes that fewer than 5% of U.S. enterprises have implemented comprehensive AI strategies, and global AI adoption is only now gaining momentum. “While the bears will continue to yell AI Bubble, we continue to point to this tech cap-ex supercycle that is driving this 4th Industrial Revolution,” Ives wrote in a note to investors. He predicts trillions of dollars in spending over the coming years will sustain the technology sector’s bull market despite broader market concerns.
OpenAI’s survival likely depends on a successful initial public offering, with the company hoping to raise as much as $1 trillion. However, this would require maintaining and exceeding that valuation indefinitely — a daunting prospect given the structural challenges facing its business.
The implications extend beyond OpenAI. The AI sector’s mounting debt, coupled with warnings from experts about an AI bubble propping up the broader U.S. economy, raises questions about systemic risk. If the industry’s poster child cannot achieve sustainability despite raising hundreds of billions, the viability of the entire generative AI ecosystem comes into doubt.
When confronted about the company’s $1.4 trillion in spending commitments against just $13 billion in revenues, Altman responded testily, telling investor Brad Gerstner: “If you want to sell your shares, I’ll find you a buyer. Enough.”
