A stark divide has emerged in the narrative surrounding OpenAI’s future. On one side, CEO Sam Altman projects boundless growth, claiming annualised revenue will exceed $20 billion this year and swell to “hundreds of billions” by 2030. On the other side, skeptics are pointing to fresh data that suggests the AI gold rush may already be slowing down, making the company’s $1.4 trillion spending plan look increasingly perilous.
The skeptics’ case is led by figures like Carl Benedikt Frey, an associate professor at Oxford University. He highlights recent U.S. Census Bureau reports showing that AI adoption among companies with over 250 employees has actually been declining. This data suggests that the initial corporate enthusiasm for AI tools may be waning as businesses “feel they are not quite getting what they hoped for.”
This trend, if it holds, directly contradicts the core of Altman’s thesis. Frey is openly skeptical of OpenAI’s ability to hit the massive revenue milestones it needs, such as a hinted-at $100 billion by 2027, “without new breakthroughs.” The entire $1.4 trillion bet is predicated on the idea that demand will not only continue but accelerate dramatically.
OpenAI vigorously disputes this narrative. The company claims it is seeing “accelerating business adoption,” stating its corporate ChatGPT version has grown nine times year-over-year. It points to its 800 million weekly users and 1 million business customers across banking, life sciences, and manufacturing as proof that demand is real and growing.
This is the central conflict: Is the slowing adoption data a blip, or is it the first sign that AI’s utility is not living up to the hype? Altman is betting his company that it’s the former. He insists the “risk of… not having enough computing power is more significant” than the risk of overspending, confident that future, more powerful models will unlock the demand needed to pay the trillion-dollar bill.