OpenAI May Be Doomed, but Not the AI Business

Wall Street may be fretting about an AI bubble, but one analyst argues the real risk isn't that artificial intelligence fizzles—it's that one of its biggest names runs out of cash first. In a New York Times opinion piece, Sebastian Mallaby contends that AI's commercial potential looks solid, citing studies that, taken together, suggest companies are already seeing meaningful returns on AI investments just a few years after ChatGPT's debut. As models get better and users more skilled, he argues, the benefits should compound. The catch: most of the billion-plus people using generative AI each month aren't paying, and free, high-quality options make it easy to switch providers.

Mallaby, a senior fellow at the Council on Foreign Relations, predicts that "stickiness" will eventually arrive, when AI agents know users so intimately that leaving them feels costly and inconvenient. "When that happens, abandoning a model might feel like a divorce—doable, but unpleasant," he writes. At that point, he says, AI firms could make serious money through subscriptions, advertising, shopping services, entertainment, and more—essentially becoming the main gateway to the internet for hundreds of millions. The question is whether independent players like OpenAI can survive long enough to reach that phase.

Unlike earlier software giants, generative AI outfits burn enormous amounts of cash, thanks to expensive computing needs. Even after Sam Altman's record-shattering $40 billion private fundraising round—bigger than any IPO haul in history—OpenAI is reportedly on track to lose billions annually, with tens of billions in projected future burn. Mallaby notes the company has also signaled plans for up to $1.4 trillion in data-center-related spending commitments.

His bet: OpenAI will run out of money within 18 months and likely be absorbed by a deep-pocketed giant such as Microsoft or Amazon, resulting in losses for OpenAI investors and forcing chipmakers and data center builders to find new customers. "Social media pundits would report every detail, and frazzled investors may dump the whole AI sector," he writes. "But an OpenAI failure wouldn't be an indictment of AI. It would be merely the end of the most hype-driven builder of it."