Analysis
Sam Altman told Fortune in an interview published September 12 that OpenAI will not go public in 2026, saying a listing this year would be an "ill-advised moment" given the current state of AI safety debate. "I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public, and we don't feel pressure on that," Altman told Fortune. He added the company still has "a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together."
From confidential filing to a public 2027 target
The delay isn't a surprise reversal so much as a confirmation of a decision OpenAI's own CFO signaled weeks earlier. Sarah Friar told employees on August 19 that OpenAI would target 2027 for a public offering, leaving the door open to moving sooner if business conditions improved, per Bloomberg's earlier reporting. Friar cited market volatility and weak recent tech-IPO performance -- SpaceX opened at $150 a share on its June 12 debut, hit a day-one valuation near $1.77 trillion, then reversed roughly 32% from its peak within two weeks. That kind of post-IPO volatility on the single largest listing in history is a live argument against rushing OpenAI's own debut before its financials and safety story are fully buttoned up.
“## From confidential filing to a public 2027 target The delay isn't a surprise reversal so much as a confirmation of a decision OpenAI's own CFO signaled weeks earlier.”
The number Altman won't compromise on
OpenAI's advisers reportedly presented Altman with a binary choice: list sooner at a lower valuation, or wait for the company to justify a $1 trillion mark. Altman has treated the trillion-dollar target as close to non-negotiable. OpenAI's most recent private valuation, $852 billion in March, held steady through an August buyback in which the company repurchased $7 billion of employee shares at that same mark -- meaning OpenAI itself isn't yet marking its own stock above $852 billion, even as it holds out publicly for $1 trillion at IPO.
Anthropic, not OpenAI, may list first
The competitive wrinkle is that Pulse's lead story today covers Anthropic CEO Dario Amodei's own safety-pacing essay, published the same day as this interview -- and Anthropic is separately reported to still be targeting an IPO as soon as October at a valuation approaching $2 trillion, roughly double OpenAI's ask, built on an annualized revenue run rate that grew explosively over the past year. If Anthropic lists first, OpenAI -- the company that built ChatGPT and still commands a larger absolute revenue base against Anthropic's newer growth curve -- would cede first-mover status in public AI markets to a rival it has out-revenued for most of the last three years.
What the headline misses
Framing this purely as a safety decision understates how much of the delay is also financial positioning: a 2027 listing gives OpenAI more time to grow into a $1 trillion number its own internal buyback hasn't yet validated, and more distance from a volatile post-SpaceX IPO market. Altman's safety framing and Friar's valuation framing aren't contradictory, but treating this as purely a principled stand ignores that OpenAI's own numbers don't yet support the price it's holding out for.
What to track next: whether Anthropic's S-1 actually prices before November as currently reported, whether OpenAI's ARR closes the gap with Anthropic's growth rate over the next two quarters, and whether OpenAI's promised independent-evaluator commitment -- made the same day as this interview -- ships with real, verifiable access before 2027 arrives.
