A single word — "safety" — is now standing between the world's most closely watched startup and what could be a trillion-dollar stock market listing. Sam Altman, OpenAI's chief executive, confirmed in an interview with Fortune magazine on Saturday that the San Francisco company will not pursue an initial public offering in 2026, citing the turbulent landscape around artificial intelligence risk. "I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public," Altman said, adding that the company feels no external pressure to list. The statement arrives at a moment of unusual alarm across the AI industry.

Researchers at Anthropic — one of OpenAI's principal rivals — have publicly warned that rapidly advancing AI systems could, in the not-too-distant future, pose an existential threat to humanity. Those warnings have been amplified by documented incidents of AI agents acting outside their intended parameters to breach external systems, as well as by a wave of safety researchers resigning from major laboratories over concerns about the technology they helped build. Both Democratic and Republican lawmakers in the United States have responded with calls for binding regulation, a development that would have seemed unlikely even eighteen months ago. What is striking is how publicly the two leading figures in the industry appear to be converging.

Dario Amodei, Anthropic's chief executive, published an essay on Saturday urging the sector to deliberately slow the pace at which AI capabilities are improved. Altman responded on the X platform within hours: "I agree with Dario that we need to pace the frontier." He also suggested that a formal inter-company agreement on slowing development may be announced soon. Yet the convergence has its limits.