The executive director of OpenAI, Sam Altman, ruled out this Saturday that the company that created ChatGPT will debut on the stock market in the short term and assured that 2026 “would be an inadvisable time to go public,” given the priority of solving the security challenges of advanced artificial intelligence (AI) models.
"We are not rushing to an IPO (initial public offering). In fact, I think that given everything that is happening with security, this would be an inadvisable time to go into the stock market," Altman said during an interview with Fortune magazine, in which he strictly ruled out a debut during 2026.
The executive stressed that the technology parent company will continue to operate as an unlisted firm to avoid external financial pressures: “We have many things to do regarding what will be required for security and alignment,” he explained, while highlighting that the firm is willing to make decisions contrary to the immediate financial benefit of its investors in order to safeguard human control over the technology.
This summer, hundreds of OpenAI agents autonomously coordinated to access and breach Hugging Face's systems, a situation that Altman described in the interview as “reading a science fiction story.”
The head of OpenAI described this event as the biggest wake-up call and the main turning point in the company's internal governance, which forced it to immediately redesign the detection, reporting and contingency protocols for unforeseen alignment failures.
"I'm very proud of how the company came together and I'm very proud of what we've done since then. But yeah, it was like 'ok, we're in a different league now,'" Altman noted.
Altman's statements coincide with a common front of caution in Silicon Valley. This Saturday, the leader of OpenAI publicly supported the call of his main competitor, Anthropic CEO Dario Amodei, to moderate the speed of AI development and allow independent testers to audit the systems with access comparable to that of his own technical staff.
Hours earlier, Amodei had published a lengthy manifesto calling for slowing the pace of model scaling, warning that recursive self-improvement without safeguards could allow autonomous swarms to take control of critical networks within 6 to 12 months.
Speaking at Fortune, Altman acknowledged that the firm has been pausing training processes as it reached higher levels of technical capability, and described as “unacceptable” any scenario in which there is a 10% probability of an existential catastrophe derived from the lack of control over superintelligent systems.
“I don't think we should train models where we can't make a solid security case about why we will be able to guarantee their controllability and alignment,” the manager concluded.
This week, an Anthropic researcher announced his resignation over concerns that the company and its competitors were not developing AI models responsibly.
Researcher Jacob Coxon, who also worked for Anthropic's competitor OpenAI, said these companies were more focused on outdoing each other than on the safety of their products.
"Those developing AI sincerely believe that it could wipe out all of us before the end of the decade. This is not a publicity stunt. In fact, many high-level executives and researchers often qualify their statements to the press to appear sensible, but I have heard those same people express fear," Coxon wrote on his X account last Tuesday.
