As investors weigh the long-term prospects of an OpenAI IPO, pressure is mounting on the San Francisco company to prove it can turn scale into sustainable profit.
Summary
Staggering valuation, huge infrastructure bill
The developer of ChatGPT now ranks among the world’s biggest startups, with a valuation of $850bn (£645bn). However, that eye-watering figure sits alongside an equally dramatic infrastructure plan, as OpenAI is reportedly set to spend $600bn on datacentres and chips by 2030, down from an initial estimate of $1.4tn.
Despite that reduced capex blueprint, the company remains far from profitability. If current trends persist, analysts warn it could burn through roughly half a trillion dollars by the end of the decade. Moreover, comparisons with earlier tech disrupters highlight the gap: Uber, for instance, spent about $30bn before finally crossing into the black, a fraction of OpenAI’s projected outlays.
Retrenchment ahead of a potential listing
Signs of strategic retrenchment have accelerated as a possible flotation, tentatively flagged for the end of this year, moves closer. Over the past month, OpenAI has shuttered three initiatives and seen another exposed as underwhelming. That said, executives and backers argue that narrowing the focus is essential if the company is to convince public markets it has a coherent path to cash generation.
In early March, OpenAI scrapped Instant Checkout, a project that allowed users to shop for goods directly inside ChatGPT. The move followed a five-month trial that underscored how difficult it is to build a robust commerce platform from scratch. An analyst at Enders noted that, like many of the company’s early launches, the feature felt more like a public demo than a serious push into online retail.
Last week, the company also axed Sora, its video-generation platform, unravelling a $1bn content-licensing agreement with Disney that was meant to “unlock new possibilities in imaginative storytelling”. The project had become a money pit for OpenAI; moreover, it was an awkward episode for Disney, which reportedly learned Sora was being canned only about an hour before the announcement went public.
In the same period, OpenAI finally pulled the plug on its repeatedly delayed erotic chatbot initiative. The plan, unveiled last year, aimed to “treat adult users like adults” by enabling sexually explicit conversations with ChatGPT. Yet with regulatory scrutiny on online safety rising, the Enders analyst said such a rollout would have been “ridiculously risky”, and a likely nightmare for both product safety and public relations.
Discipline demanded by would-be shareholders
Optimists frame these cancellations as a deliberate effort to trim the fat ahead of an openai ipo in a fiercely competitive market, where rival Anthropic and its Claude chatbot are steadily gaining traction among enterprise users. OpenAI, the Enders analyst argued, is now under “serious pressure to show strategic discipline” after casting its net too widely.
A managing director at Deutsche Bank Research Institute similarly suggested that the company is taking the right steps if reports of a potential $1tn flotation are accurate. That prospective valuation sits against annualised revenue of $25bn, a figure OpenAI is said to have reached in early March based on short-term performance. Moreover, a broader pool of public investors would almost certainly demand clear evidence that such revenue can grow in a durable, predictable way.
“If OpenAI is moving to an IPO and seeking a wider pool of investors, those investors are going to want to see real evidence of strong, sustainable revenue growth over the years to come,” the Deutsche Bank executive said. “By focusing its business model in this way, OpenAI is probably aiming for that growth in the best way possible.” He added that the company appears to have abandoned an “everything” approach in favour of a more tightly defined core.
There had long been concern, the Deutsche Bank director noted, about the lack of clear, repeatable ways to monetise what is arguably the leading consumer AI brand. However, he now sees OpenAI making the hard choices needed to solidify monetisation. “Many investors may say this is the best news they have heard from OpenAI in months,” he said.
Subscriptions and enterprise deals at the core
Despite the turbulence, the flagship product at the heart of the AI boom remains hugely popular. ChatGPT has passed 900 million weekly active users and boasts more than 50 million paying subscribers. Those subscriptions account for about 75% of OpenAI’s income, and the company also sells corporate-grade ChatGPT offerings while letting firms and startups build tailored products on top of its models.
Even so, analysts argue that OpenAI should have imposed financial discipline earlier, given its habit of burning through billions of dollars each month on experiments that often fail to mature into real businesses. One columnist recently dubbed it “the most distracted company in technology” after Instant Checkout was abandoned. Moreover, frequent product announcements that never reach scale have added to investor unease.
“We have seen so many consumer product launches, promising to disrupt the browser, online commerce, content creation, search,” the Enders analyst said. “Actually focusing your strategy and executing on a product that people will want to use and, crucially, be willing to pay for in some real form, is the harder challenge.” That assessment goes to the heart of the company’s profitability challenges.
Ads emerge as a new revenue test
Amid the cutbacks, OpenAI last week reported what looked like a rare unambiguous win: its trial of advertising in ChatGPT has generated $100m in annualised revenue, implying around $12m earned over six weeks. Because the service holds rich data about user intent and behaviour, the idea of highly targeted promotions appears, in principle, to be a powerful monetisation lever.
The Forrester analyst tracking the pilot said the ai advertising trial results were “better than expected”. However, she stressed that this does not mean the company is close to building a mature ad business. Moreover, the technical and regulatory demands of delivering personalised ads at scale could be significant, especially in an environment of tightening privacy rules.
The Enders analyst also warned that getting advertising right would take far more work. “It could very quickly begin to feel creepy and risk a user backlash and privacy concerns,” she said. On the other hand, she added, ads inside ChatGPT would not drive meaningful revenue if they remained little more than “a glorified banner ad below the answers” without deeper integration or intelligent targeting.
Compute constraints and capital intensity
Beyond business-model questions, OpenAI faces an enormous operational challenge: the cost and scarcity of the computing power needed to run its models. The company has acknowledged that “compute” is in short supply globally and says it is prioritising how and where it invests. That said, its long-term infrastructure bill remains one of the most striking aspects of its strategy.
“With user demand outstripping supply, compute is the critical resource when it comes to AI,” an OpenAI spokesperson said. The company argues that locking in long-term compute through its infrastructure strategy, while ruthlessly allocating that capacity to the highest-value uses, is essential to limit its staggering cash burn.
According to the spokesperson, OpenAI is concentrating compute on three main areas: advancing frontier research, expanding its base of more than 900m users worldwide, and powering enterprise use cases. Moreover, executives say that as they secure more large-scale compute, this disciplined allocation allows the company to innovate faster and serve both enterprises and developers more efficiently.
Can OpenAI justify a trillion-dollar float?
For now, OpenAI sits at a crossroads. It commands vast cultural and technological influence, yet must still prove that its breakthrough products can support a business model strong enough to justify a potential $1tn market debut. Investors, eyeing a prospective listing later this year, are waiting to see whether management can convert surging adoption into durable profits before the stock market decides how much that promise is really worth.
