OpenAI has committed to spending over $1 trillion on AI infrastructure, including 26 gigawatts of compute capacity from Nvidia, AMD, and Oracle, according to Financial Times calculations. Yet the company's revenue base remains razor-thin: only about 5% of its 800 million ChatGPT users pay for subscriptions, and losses in the first half of the year reached $8 billion.
The gap between spending and income has become a central concern not just for OpenAI but for the broader economy, as capital expenditures on AI have contributed more to US economic growth than consumer spending as of August, per the Wall Street Journal. This has amplified fears of an AI bubble, where a sudden downturn could have ripple effects across markets.
Sam Altman, OpenAI's CEO, has acknowledged the need for profitability but has not made it a top priority. In a recent press briefing, he said breaking even is “not in my top ten concerns, but we obviously someday have to be very profitable.” The company is exploring new revenue streams, including advertising, monetizing its text-to-video generator Sora, and venturing into online shopping. It is also developing a personal device with former Apple designer Jony Ive.
However, these initiatives face significant hurdles. Sora, which is resource-intensive, could become a major expense rather than a reliable income source. The personal device market has seen failures like the Humane AI pin, suggesting that even with Ive's involvement, success is uncertain.
In an attempt to boost engagement, Altman reversed a previous stance, announcing that ChatGPT would allow “mature apps,” despite earlier claims that the platform hosted no “sexbots.” This move has raised eyebrows but reflects the pressure to increase user activity.
Why the Revenue Gap Matters
The core issue is that OpenAI's subscription model—$20 per month for standard access and $200 for a pro tier—has not convinced the vast majority of users to pay. With only 40 million paying subscribers, the revenue generated is insufficient to cover operational costs, let alone the massive infrastructure investments.
OpenAI is also taking on more debt and relying on partners like Oracle to share the financial burden of infrastructure spending. This gives the company “time to build the business,” as one senior executive told the FT, but insiders admit there is no cohesive long-term plan. The same executive noted, “[Investors] expect you to have a five-year model,” but added, “right now I’d say there’s lots of fuzz on the horizon.”
Analysts have also flagged circular deals, such as OpenAI using investment money from Nvidia to purchase hardware from the same company, which heighten concerns about the sustainability of the AI boom.
The stakes are high. If OpenAI's spending continues without a corresponding revenue surge, the fallout could extend beyond the company, affecting the wider economy that has come to rely on AI's growth. Whether Altman's strategies will close the gap remains an open question, but the current trajectory suggests that the road to profitability is steep.
