If OpenAI intends to go public this year, it must subject its business model to rigorous scrutiny. As an American enterprise emblematic of the AI industry's boom, the company has achieved widespread renown; yet, it has consistently struggled to resolve its profitability challenges, and its plans for an IPO have sparked market concerns regarding a potential stock market bubble.
OpenAI—the developer behind ChatGPT—currently ranks as one of the world's largest startups, boasting a valuation of $850 billion (£645 billion). Concurrently, reports suggest the company plans to invest $600 billion in infrastructure by 2030—an outlay covering everything from data centers to the chips required to power its AI models. Although this investment scale represents a reduction from the initially projected $1.4 trillion, it remains a staggering sum.
Even with these scaled-back spending plans, OpenAI still faces a long and arduous journey toward achieving profitability. Indeed, if current trends persist, the company’s cumulative cash burn could reach $500 billion by the end of the decade. While proponents might cite Uber as a precedent—noting that the ride-hailing giant also invested billions before turning a profit—it is crucial to recognize that Uber’s total investment amounted to a mere $30 billion, a figure dwarfed by OpenAI’s projected $600 billion expenditure.
Under the leadership of CEO Sam Altman, OpenAI is accelerating its decision-making pace—partly in anticipation of an impending market correction, and partly to align with its plans for a public listing by year-end. Within just the past month, OpenAI has proactively abandoned three distinct business initiatives, while a fourth has been confirmed to face bleak prospects.
In early March, OpenAI formally scrapped its "Instant Checkout" initiative—a project designed to enable consumers to complete purchases directly within the ChatGPT interface. The company had previously conducted a five-month pilot run of the program, only to discover that building a successful e-commerce platform proved far more challenging than anticipated. Niamh Burns, an analyst at Enders, observed: "Much like many of OpenAI’s earlier projects, this endeavor appeared to be more of a public demonstration of technical capabilities than a sustained, strategic effort to establish a viable e-commerce business."
Just last week, OpenAI not only shelved its video-generation platform, Sora, but also terminated a $1 billion partnership agreement with Disney. Under the terms of the agreement, Disney had planned to license generative content from OpenAI to "unlock new possibilities for imaginative storytelling." For OpenAI, this represents a strategic move to scale back operations, given that funding the Sora project had become a financial black hole; for Disney, however, this turn of events is highly awkward—reports suggest the company learned of the decision to shut down the Sora platform just one hour before the news was made public.
Furthermore, last week OpenAI also halted its previously announced project for an erotic chatbot. First unveiled last year, the project was conceived with the intention of "treating adult users like adults," allowing them to engage in sexually suggestive conversations with ChatGPT; however, its launch date had been repeatedly postponed. "This would have been an incredibly risky launch," noted Burns, particularly in the current climate of increasingly stringent cybersecurity scrutiny. "From both a product safety and a public relations standpoint, it would have been an absolute nightmare."
On a more optimistic note, these moves suggest that OpenAI is streamlining its business structure in preparation for an initial public offering (IPO). The AI market is currently fiercely competitive, and Anthropic—the creator of the Claude chatbot—is steadily winning over more enterprise clients, thereby exerting significant competitive pressure on OpenAI. Burns pointed out that OpenAI "faces immense pressure to demonstrate strategic discipline"; having previously "cast its net too wide," the current scaling-back represents a necessary course correction.
Adrian Cox, a Managing Director at the Deutsche Bank Research Institute, stated that if reports are accurate—specifically that OpenAI is gearing up for a public listing with a projected valuation of $1 trillion—then the streamlining measures the company is currently undertaking are the right ones. By way of comparison, OpenAI’s annualized revenue (based on short-term performance projections) has already reached $25 billion—a milestone reportedly achieved in early March.
"If OpenAI plans to go public and seek a broader investor base, those investors will certainly want to see tangible evidence that the company can deliver strong, sustainable revenue growth over the coming years," Cox observed. "By adjusting its business model in this manner, OpenAI may well be pursuing the optimal strategy to achieve that objective."
He further added that OpenAI appears to have ceased attempting to go head-to-head with competitors employing an "all-encompassing" business model; instead, it has opted to narrow its operational scope and refocus its efforts on core business areas. "People have long worried that OpenAI—currently the leading consumer AI brand—lacks a clear path to profitability," said Cox. "Now, it appears OpenAI is making tough choices to better position itself for future profitability. Many investors might say this is the best news they’ve heard from OpenAI in months."
As the flagship product of both OpenAI and the broader AI boom, ChatGPT continues to enjoy immense popularity. The product currently boasts over 900 million weekly active users and more than 50 million paying subscribers. OpenAI derives the majority of its revenue—75% of the total—from these subscription services; additionally, the company supplements its income by offering an enterprise version of ChatGPT to businesses and by allowing various companies and startups to utilize its AI models to develop their own products.
However, analysts generally agree that OpenAI should have demonstrated greater strategic discipline much earlier—particularly given that it has been spending billions of dollars monthly on experiments that ultimately yielded minimal results. Following the failure of its "instant checkout" project, a *Forbes* columnist went so far as to label OpenAI "the most distracted company in tech."
"We’ve seen a flurry of consumer product launches, all promising to disrupt web browsers, e-commerce, content creation, and search," said Burns. "In reality, the far more difficult challenge lies in focusing your strategy on developing a product that people actually want to use—and, crucially, are willing to pay for in some tangible form."
Last week, amidst a backdrop of internal turmoil, OpenAI released a seemingly positive piece of news: its pilot advertising program for ChatGPT has reached an annualized revenue run rate of $100 million—meaning the pilot generated approximately $12 million in just six weeks. This could potentially open up a new avenue for profitability for OpenAI; after all, given its deep understanding of its user base, ChatGPT possesses the potential to deliver highly targeted advertisements.
Even so, Burns cautioned that—much like every other project the company has attempted in the past—building a successful advertising business will likely require significant effort. "It could very quickly start to feel 'creepy,' potentially triggering a strong backlash from users and raising serious privacy concerns."
On the other hand, she noted that if the advertisements within ChatGPT amount to nothing more than "premium banner ads displayed beneath the answers"—without any form of targeted delivery—then such advertising is unlikely to generate substantial business revenue. Forrester analyst Nikhil Lai stated that the results of the recent advertising pilot were "better than expected," though this does not mean OpenAI is close to achieving its goal of profitability through advertising.
Lai noted that it could take OpenAI "several years to reach that goal—if they ever do," adding: "They have a lot of work to do, and they need to make a lot of changes."
The company, which holds the world's most high-profile AI technology, must quickly identify a viable business model while curbing its unsustainable cash burn. Investors are eagerly awaiting clarity regarding OpenAI's future trajectory.
An OpenAI spokesperson stated that the company is currently prioritizing infrastructure investment, citing a shortage of the infrastructure required to run its AI models (known as "compute").
"Compute is a critical resource in the AI space, as user demand currently outstrips supply," the spokesperson said. "In addition to securing our long-term compute needs through our infrastructure strategy, we are uncompromisingly prioritizing the allocation of these resources to areas that generate the greatest long-term economic value: advancing frontier research, expanding our global user base of over 900 million, and supporting enterprise use cases."
"As we continue to secure access to increasingly large-scale compute resources, this rigorous approach to allocation enables us to grow, innovate faster, and serve enterprises and developers more efficiently."
