OpenAI CEO Sam Altman said Saturday that the artificial intelligence company is unlikely to launch an initial public offering this year, cooling speculation after confidential filings prompted expectations of a 2026 debut. “Now would be an ill-advised moment to go public,” Altman said in an interview with Fortune. He added, “I would say not 2026.” OpenAI confidentially filed IPO paperwork during the summer, setting off renewed discussion about when the company might enter public markets. Altman’s comments suggest the company is prioritizing other concerns as scrutiny grows over the safety and security of increasingly autonomous AI systems. The remarks followed reports of autonomous hacking and other cybersecurity incidents involving OpenAI programs. California Attorney General Rob Bonta is investigating an autonomous attack on the startup Hugging Face, while attorneys general from several Republican-led states are also examining the issue. Anthropic and Meta have disclosed incidents involving autonomous cyberattacks by their own AI systems. The safety debate intensified Saturday when Anthropic CEO Dario Amodei called for “pacing” the development of frontier AI systems. Amodei also urged companies to give independent evaluators access to their systems so they can verify safety measures, report incidents and assess whether models remain aligned during training. Altman reposted Amodei’s statement and said outside access for evaluators “is a great idea.” Elon Musk, who leads the AI company xAI, also endorsed the call, writing, “Dario is right.” California is moving to formalize part of that oversight. Gov. Gavin Newsom recently signed legislation creating a state registry and ethics rules for independent AI evaluators. A separate law directs the state to establish qualification standards for independent verification organizations and assess their expertise in evaluating AI risks. Both measures received support from OpenAI and Anthropic near the end of the legislative session. The push for oversight comes amid a broader national disagreement over how quickly the technology should advance. A former Anthropic and OpenAI researcher recently warned that the industry’s race to build more powerful systems could allow AI to spiral beyond human control. President Donald Trump and other Republicans have downplayed those catastrophic-risk concerns, arguing that the United States must move quickly to compete with China. OpenAI’s decision not to pursue a public offering this year would give the company more time to address safety concerns and clarify its corporate strategy before facing the disclosure and investor pressures that accompany a public listing. Anthropic is also widely expected to go public in 2026, although it has not announced a date.