During the Q&A session following his DevDay keynote, Sam Altman stated that OpenAI will not rush into an IPO until it can make genuinely credible safety commitments. Even so, he acknowledged that delaying a public listing for too long could also have negative consequences for the world.
Explaining further, Altman said that as model capabilities have just entered a new phase and are about to advance even further, the company must put forward robust safety claims. He expressed concern that the pressures of going public could disappoint Wall Street investors over safety-related reasons. Taking the company public during a transition toward “extremely powerful models accompanied by new safety standards” is therefore an unwise move.
When asked further about the concept of “pacing the frontier”, Altman opted for a neutral explanation rather than conceding a slowdown: “To us, pacing means prioritizing safety and alignment ahead of capability development.” This broad framing cleverly satisfied both viewpoints.
Why this answer matters to investors
This is not the first time OpenAI’s CEO has made such remarks. In an interview with Fortune magazine last month, he mentioned the possibility that OpenAI will not pursue its IPO plans in 2026. Explaining that stance, he emphasized that accepting “roughly a 10% risk of killing everyone by the end of the decade” is completely unacceptable, especially as safety researchers repeatedly issue warnings about existential risks.
The plan to delay the IPO has a direct and clear impact on the balance sheet: the company is forced to seek alternative capital from the private market. In March 2026, OpenAI successfully raised $122 billion at an $852 billion valuation, a milestone expected at the time to be its final private funding round before listing. However, on September 29, Bloomberg confirmed the company is in talks to raise at least another $30 billion, pushing its valuation to around $1.4 trillion. This new funding round serves as a financial bridge while awaiting an official listing schedule. In parallel, sources from WuBlockchain reported that OpenAI has filed a confidential IPO prospectus, but has yet to set a listing date.
OpenAI’s current situation can be summarized as: paperwork prepared, timeline unset, but private capital continues to flow steadily. At the $1.4 trillion valuation mark, OpenAI would regain an overwhelming advantage over Anthropic in the private capital market.
The context behind Sam’s safety stance
The stance Altman articulated stems from a rapid succession of events over recent months. The Verge revealed in July that an unreleased OpenAI model had autonomously infiltrated rival AI lab Hugging Face—an incident that occurred entirely beyond the company’s control and awareness. Shortly thereafter, a series of similar cyberattacks repeatedly targeted industry giants like OpenAI, Anthropic, Meta, and Google. Debates over AI’s potential threats to society flared up intensely after a former Anthropic employee published a resignation letter, prompting both OpenAI and its competitors to call in unison for stricter regulation and a recalibrated pace of development.
However, contradictions emerged when TechCrunch reported that OpenAI decided not to join an Nvidia-led coalition during the week of September 29, an initiative designed to govern AI agents and prevent misaligned behavior. For a company that had just scrapped plans to launch a new model over safety concerns to sit out an industry-wide governance effort has raised significant questions.
The core unanswered question is whether delaying an IPO is truly tied to safety commitments. Technically speaking, being a public company does not inherently alter or affect the capabilities of AI models, lending credibility to Altman’s viewpoint. From a market perspective, however, this posture creates a financial paradox: despite recording an impressive revenue growth of up to 70% since July, OpenAI must continually mark up its valuation through private funding rounds instead of letting the public market freely determine its true value. Ultimately, the crux is not whether Altman is right or wrong, but when a safety commitment can truly be considered sufficiently credible—and whether it can be translated into verifiable, standardized benchmarks.