A source familiar with the matter has revealed that OpenAI has completed a transaction to purchase approximately $7 billion worth of shares from its employees, paving the way for a potential initial public offering. Two insiders confirmed that the company bought back shares from both current and former staff in this tender offer, departing from past practices where external investors like Thrive Capital and SoftBank Group were invited to purchase employee-held stock. This deal values the startup at $852 billion, matching its valuation from the most recent funding round.

The share buyback initiative was set in motion after OpenAI closed a record-breaking $122 billion funding round in March. The transaction is designed to ease near-term liquidity pressures, allowing employees to sell part of their holdings and realize gains ahead of a possible large-scale IPO. Secondary market stock sales have become a key part of OpenAI's pre-IPO strategy. In October of last year, the company completed a $6.6 billion tender offer at a valuation of $500 billion, and it also executed a $1.5 billion buyback in 2024. Such transactions are increasingly common in Silicon Valley, with companies like Stripe, Databricks, and SpaceX offering employee share sales or secondary market deals to provide liquidity despite delaying their own listings.

Citing data from Carta, The Information reported that current and former employees at private companies sold about $1.7 billion in shares in 2025, exceeding the combined total for 2024 and 2023 by more than $200 million. The longer a company delays its IPO, the more stock options and restricted stock units become a burden in terms of taxes and cash flow. Javier Avalos, CEO of private market data platform Caplight, noted that OpenAI employees have been fully vested and holding shares for some time, creating pressure on the company to offer liquidity. Company-led share buybacks are the primary avenue for employees to cash out equity before an IPO, though some also sell through individual secondary transactions.

However, employee share sales only address internal liquidity needs and do not resolve the capital demands of frontier model companies. The Information pointed out that for OpenAI and its rival Anthropic, an IPO could raise tens of billions of dollars to fund the training and operation of AI models. Both companies are expected to spend hundreds of billions on computing services over the next decade. Not all employees are eager to sell before the IPO. The Information noted that Anthropic's employee share sale earlier this year fell short of the previously reported $5 billion to $6 billion in investor buying interest, partly because some employees believed they would get a better deal by waiting until after the listing.

Meanwhile, OpenAI and Anthropic have been cracking down on unauthorized share transfers, such as those conducted through special purpose vehicles. For AI companies preparing to go public or raise large sums, controlling the order of equity transactions is a necessary step before entering public markets. Reports also indicate that OpenAI has significantly stepped back from its initial ambition of an IPO as early as this fall, now clearly leaning toward delaying the timeline to 2027. This postponement stems from a clash between CEO Sam Altman's insistence on a trillion-dollar valuation floor and harsh market realities. The volatile stock performance of SpaceX after its listing has served as a direct psychological deterrent to OpenAI's IPO plans.

Bankers advising OpenAI on the IPO have explicitly warned that recent turbulence in tech stocks and the sharp decline in SpaceX's share price post-listing could severely dampen retail investor enthusiasm for OpenAI's offering. One source revealed that over the past week, advisors told the company that retail investors may lack excitement for the stock. OpenAI's valuation predicament is the core issue behind the delay. In March 2026, the company raised $122 billion in funding, achieving a post-money valuation of $852 billion and making it the world's most valuable private tech company. Yet this falls short of Altman's expectations. According to insiders, Altman has been pushing his team of bankers and lawyers to find ways to boost the IPO valuation to $1 trillion. The advisory team presented two options: delay the IPO to 2027, allowing market conditions to improve and financial performance to align with the trillion-dollar target, or list by late 2026 but accept a lower valuation. When presented with this choice, Altman reportedly dismissed any valuation below $1 trillion as "unworkable."

OpenAI's financial health is also testing investor patience. The company reported a net loss of $38.5 billion last year, driven by massive spending on computing infrastructure, research and development, and corporate restructuring. The Information reported that in the first quarter of 2026, OpenAI burned through $3.7 billion in cash, exceeding half of its $5.7 billion in revenue for the period. The company expects to invest $600 billion in computing and hardware by 2030. In recent months, some major investors have privately expressed concern about OpenAI's cash burn rate relative to its growth, while others have hedged their bets by investing in Anthropic, according to insiders.

OpenAI's delay is not an isolated case. Analysts suggest that large language model companies initially expected to go public in the second half of 2026 may now be pushed to the first half of 2027 due to declining market risk appetite and liquidity uncertainties. The adjustment to the IPO timeline signals a significant postponement from the earlier market expectation of this fall. The most straightforward reading of the 2027 timeline is that OpenAI can afford to wait. By delaying, the company can continue expanding usage, refining pricing, and developing a more stable mix of consumer products, enterprise tools, and infrastructure partnerships before facing the quarterly discipline of public markets. According to the latest reports, OpenAI's annualized recurring revenue in July exceeded its total revenue for the entire second quarter.