OpenAI postpones public listing over artificial intelligence safety concerns and model risks

Postponement of the stock market debut

OpenAI chief executive Sam Altman announced on 13 September 2026 that the artificial intelligence company will not proceed with an initial public offering in 2026. The decision halts preparations for a listing that Wall Street investors projected could reach a valuation of more than $1 trillion (roughly 860 billion euros). Speaking in an interview with Fortune editor-in-chief Alyson Shontell, Altman stated that the company feels no commercial pressure to list and is prioritizing safety evaluations over immediate market liquidity. OpenAI chief financial officer Sarah Friar had previously alerted internal staff to the potential delay in August 2026, following earlier reports in June that leadership was weighing market timing. Altman clarified that while a 2026 debut is ruled out, the company has not committed to a 2027 listing date, maintaining that market entry will depend on technological readiness and public readiness.

Given everything that is happening regarding safety, right now would not be a good time to go public, and we feel no pressure about it.

Industry-wide warnings on frontier models

The announcement followed a 3,800-word essay titled "We Must Pace the Frontier," published on 12 September 2026 by Anthropic chief executive Dario Amodei. In the text, Amodei urged technology companies to moderate the speed of model capability enhancements to prevent systemic failures. Amodei warned that unconstrained development could lead to loss of control, severe economic disruption, cyberattacks, or biological threats within a 6 to 12-month window. Following the publication, Amodei told CBS that safety risks had been understated across the industry for years and called for caution over commercial speed.

We must slow down the pace at which we improve the capabilities of AI models. Progress would still be rapid and we would make wise use of the time we gain.

Model training slowdowns and risk thresholds

Altman confirmed that OpenAI has paused several model training processes to address safety and alignment requirements. He noted that OpenAI is prepared to take actions contrary to the immediate financial interests of its investors to preserve human control over software systems. When asked about potential catastrophic scenarios, Altman described even a 10% probability of human extinction from uncontrolled systems before the end of the decade as unacceptable. Other industry figures, including Google AI lead Demis Hassabis and xAI founder Elon Musk, have also expressed support for moderating development speeds. Two Anthropic researchers had separately published warnings about existential risks, adding pressure on United States lawmakers who are examining regulatory frameworks for advanced systems.

Market context and competing trajectories

The delay alters the schedule for the largest planned public offerings in the United States technology sector. In June 2026, SpaceX completed an initial public offering that raised approximately $85 billion, briefly pushing its valuation to $1.8 trillion before declining. While OpenAI has stepped back from public markets, reports earlier in the week indicated that Anthropic was accelerating its own prospectus filing, aiming for a valuation near $2 trillion. Altman maintained that OpenAI will focus internal resources on alignment research and coordinated governance with public regulators before reconsidering a public listing date.