Sam Altman’s OpenAI stands at a crossroads few companies ever reach. In late March it closed a staggering $122 billion funding round that lifted its post-money valuation to $852 billion. Weeks later fresh figures showed the company burned $3.7 billion in the first three months of 2026 alone. That cash drain exceeded half the $5.7 billion in revenue it generated during the same period.
The numbers tell two stories at once. One of explosive demand. Another of costs that refuse to bend. And both point to the same pressure: prove the model works before public markets demand answers.
OpenAI itself announced the fundraising on its website. Revenue had climbed from $1 billion within a year of ChatGPT’s debut to $1 billion per quarter by the end of 2024. Now the company generates $2 billion per month. It grew four times faster than the internet and mobile pioneers such as Alphabet and Meta. Enterprise customers now account for more than 40 percent of sales and sit on track to match consumer revenue by year-end. Weekly active users top 900 million. Paying business users exceed 9 million.
Yet the burn rate tells its own tale. The Information obtained documents shared with shareholders that paint a stark picture. Computing, power and data-center expenses drive most of the outflow. The firm carries hundreds of billions in future spending commitments. Cash and marketable securities topped $73 billion after the March round. Still, the pace of consumption raises eyebrows among even bullish investors.
Earlier financials offer context. OpenAI reported $13.1 billion in full-year 2025 revenue according to multiple accounts including CNBC. Losses, however, reached enormous scale in prior periods. One analysis of audited figures reviewed by the Financial Times showed costs and expenses hitting $34 billion against that revenue, producing operating losses near $21 billion before adjustments pushed net losses even higher in some estimates.
But the latest quarter’s figures hit harder because they arrived after the massive capital infusion. Tripling both revenue and burn from the year-ago period signals scale without efficiency. Chips and energy alone consume vast sums. No one disputes the demand. The question is whether those dollars can ever translate into sustainable margins.
Investors lined up anyway. Amazon anchored the round with a $50 billion commitment, much of it reportedly contingent on an IPO or reaching artificial general intelligence. NVIDIA and SoftBank each contributed $30 billion. Microsoft, Andreessen Horowitz and D.E. Shaw joined as well. Roughly $3 billion came from retail investors through bank channels, a first for the company. The deal ranks as the largest private fundraising in history.
That enthusiasm reflects belief in the technology’s long-term payoff. It also reveals something about the current market. Capital chases AI exposure at almost any price. OpenAI’s valuation now sits in territory once reserved for entire sectors. Comparisons to Berkshire Hathaway surface in coverage from CoinDesk. Yet Berkshire generates actual profits.
Rivals add tension. Anthropic surged past OpenAI in valuation during May, closing a $65 billion round at $965 billion according to CNBC and The New York Times. Some reports claim Anthropic also leads in annualized revenue. The two labs trade barbs and talent. OpenAI recently hired Noam Shazeer, a Google DeepMind legend, and former Trump administration AI policy official Dean Ball to bolster its strategic posture as it prepares for public life.
Altman told staff in early June that he expects an IPO within the next year, per Reuters. The company confidentially filed its S-1 in recent weeks. CFO Sarah Friar has spoken of adopting public-company discipline. Those steps matter because the current structure, built around a nonprofit parent and complex capped-profit subsidiary, will face fresh scrutiny from regulators and ordinary shareholders.
Public markets will demand more than growth headlines. They will want a credible path to profit. So far that path remains obscured by compute costs that rise in lockstep with capability. Newer models such as GPT-5.4 drive agentic workflows and lift enterprise adoption. They also intensify the need for specialized hardware and electricity. One internal projection once eyed $1.4 trillion in spending by 2030 before being reset toward $600 billion. Even the lower figure dwarfs most corporate budgets.
Analysts differ on timing. Sacra estimates placed annualized revenue near $25 billion as of February 2026. Some forecasts see $39 billion by mid-2027, though with wide error bands. Others warn that without dramatic efficiency gains the losses will compound. High multiples invite skepticism. A 34x revenue valuation on $25 billion run-rate cash-burn numbers strikes some observers as stretched.
OpenAI counters that its trajectory outpaces every prior technology wave. ChatGPT reached a billion users faster than any product in history. The shift toward business use cases, from coding to customer support to scientific research, suggests sticky revenue. New usage analytics and spend controls rolled out in June aim to give enterprise customers better visibility and governance. Those features matter for adoption at scale.
Yet technical benchmarks reveal limits. OpenAI’s own LifeSciBench, released in June, tested frontier models on 750 real research tasks judged by PhD scientists. Top systems cleared only 36 percent. The result underscores that even leading AI still falls short on complex, real-world work. Progress continues. So do the gaps.
Policy and regulatory clouds loom too. Dean Ball’s new Strategic Futures team will tackle catastrophic risk, self-improvement, labor impacts and government relations. Those issues grow urgent as models grow more powerful. Altman has long warned about both the promise and peril of advanced AI. The company’s nonprofit roots were meant to keep safety first. The profit engine now dominates the conversation.
Recent product moves show focus. Improvements to health intelligence in ChatGPT, AI tools for diagnosing rare genetic diseases, and research on predicting model behavior before release all signal investment beyond pure language generation. Each carries heavy computational weight.
The coming year will test whether OpenAI can translate its market leadership into financial credibility. Revenue growth looks real. User numbers look real. The cash consumption looks real too. Public investors will weigh all three when shares eventually trade.
For now the company enjoys an extraordinary cash cushion and continued private-market enthusiasm. But the clock ticks louder. Altman’s message to staff carried both confidence and urgency. The next phase, he and his backers believe, will justify the extraordinary sums already committed.
Whether it does depends on costs finally bending downward. Or revenue accelerating beyond even the optimistic projections. The data so far shows both moving in the same direction. Up. Fast. The margin between them will decide if this becomes one of the great business successes or a cautionary tale about scaling artificial intelligence.
WebProNews is an iEntry Publication