On June 8, local time in the United States, OpenAI announced on its official website that it had formally confirmed the secret submission of its S-1 filing to the U.S. Securities and Exchange Commission. The AI giant, valued at over $850 billion, has finally taken a substantial step towards the public market.

This not only signifies that Wall Street may see one of the largest tech IPOs in history, but also intensifies the capital race between Sam Altman and his archrival Elon Musk, along with the up-and-coming Anthropic CEO Dario Amodei.

As per usual practice, a secret submission of an IPO filing should have been an operation conducted under strict confidentiality. Sam Altman had no intention of passively waiting for media leaks. In the OpenAI announcement, it was stated: “We recently submitted our S-1 filing confidentially. We expected it to leak, so we decided to just make it public.”

However, OpenAI also tempered the overheated expectations in the official press release, explicitly stating that the timing of the IPO has not yet been decided. “It may take some time because there are some things we want to do that might be easier as a private company. But this is a set of complex trade-offs that gives us an option: if going public ends up being the best path, we can move more quickly to the public markets.”

This statement not only demonstrates a willingness to embrace capital but also leaves room to maneuver between an unfinished mission and significant profits. This filing was made pursuant to Section 135 of the 1933 U.S. Securities Act and does not constitute an offer to sell securities.

OpenAI’s CFO Sarah Friar stated in April of this year that for a company of OpenAI’s scale, appearing, feeling, and acting like a public company is a “good norm,” but she did not comment on a specific timeline at that time. Insiders familiar with the matter revealed that OpenAI also plans to conduct a tender offer, allowing employees to sell shares at the latest post-investment valuation of $852 billion to ease recent liquidity pressures.

OpenAI has revealed its hand, essentially mirroring its “mortal enemy” Anthropic’s path. Just on June 1, Anthropic announced a secret IPO filing, surpassing OpenAI’s valuation set at the end of March at $852 billion with a valuation of $965 billion in the private market.

Meanwhile, Elon Musk’s SpaceX has already kicked off its IPO roadshow and is set to go public as soon as June 12. In its listing documents, OpenAI, Anthropic, and Google are all listed as “key competitors” in the AI space. These three companies are highly likely to be involved in some of the largest IPOs in human history.

Jeff Bernstein, Capital Markets Advisor at consulting firm Riveron, pointed out the nature of this race: “It’s a capital grab.” He suggested that letting the other party go first would take away a significant amount of available IPO capital. However, going public first does not guarantee better returns.

Take Lyft and Uber, for example. Both companies secretly filed for IPO in December 2018, with Lyft going public in March 2019 ahead of Uber, which followed two months later. But a year after going public, Lyft’s stock price dropped by around 66% from its IPO price, while Uber only fell by about 30% during the same period.

According to official data, within a year of ChatGPT’s release, the company’s annual revenue exceeded $1 billion. By the end of 2024, this number had evolved to $1 billion in quarterly revenue. Currently, OpenAI has reached a monthly revenue of $2 billion. The officials even boldly claim that its revenue growth rate is four times that of Internet and Mobile era giants like Alphabet and Meta during the same period.

On the user front, ChatGPT’s weekly active users have surpassed 9 billion, with over 50 million subscribers. Its monthly web visits and mobile sessions are six times that of the next leading AI app, with a total duration share four times that of competitors as a whole, and four times that of all other AI apps combined. Search usage has nearly tripled in a year, and an advertising pilot project achieved over $1 billion in annual recurring revenue in less than six weeks.

The enterprise market currently accounts for over 40% of the company’s revenue and is expected to be on par with the consumer business by the end of 2026. Driven by GPT-5.4, the utilization of AI workflows is experiencing explosive growth, with an API throughput surpassing 15 billion tokens per minute. Codex has exceeded 2 million weekly active users, growing fivefold in the past three months, with a month-over-month growth rate of over 70%.

Behind the impressive revenue numbers lies a cash burn black hole that is hard to ignore for any publicly traded company. OpenAI has raised over $180 billion in funding, but the cash burn during the process of ensuring computing power and building infrastructure to train and run AI models is staggering. Its shared forecasts with investors indicate that based on the tens of billions of dollars commitment signed until 2030, its cash burn rate outstrips any other publicly traded company in history.

The article compares the proliferation of AI to the electrification of rural America in the 1920s. Electricity did not transform every household overnight, but as it spread, daily life changed fundamentally, extending the night with lights, easing heavy labor with appliances, and bringing distant voices into the living room with radios. By the end of the 20th century, the average human lifespan had extended by about 23 years, and the average income had grown by about 50% inflation-adjusted, largely due to electrification and related technological advances.

“This scene is now playing out with AI,” the two authors wrote. However, they quickly point out that such a future will not arrive automatically. “Transformational technologies can concentrate power or broaden power. They can make life easier for a few or open up opportunities for many.”

The article divides OpenAI’s history into three stages. The first stage was focused on AGI research. The second stage began with connecting research to the real world, becoming a product company, deploying systems, and learning how people interact with them. The article outlined three main goals: to build an automated AI researcher, with an internal belief that by March 2028, a significant portion of research will be conducted jointly by AI systems and researchers; to accelerate economic development, ensuring benefits are widely shared; and to provide personal AGI for everyone on Earth.

Just as OpenAI filed for an IPO, the company was also brewing a transformation at the product level. On June 7, the Financial Times reported, citing sources familiar with the matter, that OpenAI was undergoing a major overhaul of ChatGPT, planning to integrate the programming tool Codex into the chat interface as a whole, and to integrate with external partner applications, reshaping it into a multitasking super portal.

However, before the product actually lands, the more pressing test still comes from the public market. OpenAI, SpaceX, and Anthropic are all seeking listings at massive valuations, with the three companies collectively possibly raising up to hundreds of billions of dollars from the public market.

Bankers have told these companies that whoever goes public first will define this nascent industry and be able to take advantage of significant funds from investors eager to invest in AI companies. Just as the noise from the lawsuit with Musk has subsided, Ultraman is leading OpenAI into a future he himself calls a “complex balance.”

The response of investors to SpaceX’s large-scale IPO, the overall health of the global economy, the company’s own financial strength, as well as unpredictable revenue growth and soaring computing costs, will all impact OpenAI’s final IPO timeline. The answers may lie in the yet-to-be-released IPO prospectus.

[BlockBeats]