OpenAI secretly started the IPO process, and Altman said "We will go public within a year," planning to initiate an employee tender offer at $687.69 per share.

The Chief Executive Officer of OpenAI, Sam Altman, revealed to employees through a Slack message that he anticipates the company to go public within "the next year". He added, "Many factors could lead to an earlier or delayed listing within this timeframe, but filing now gives us the option to go public when we choose." Additionally, on the same Monday, OpenAI disclosed that they have secretly submitted a draft IPO prospectus to the U.S. Securities and Exchange Commission (SEC). Reports indicate that Altman also informed employees that the company is preparing to launch a tender offer at the current price of $687.69 per share. He also hinted that delaying the IPO might be more advantageous if artificial intelligence can achieve "Recursive Self-Improvement (RSI)". He stated, "The potential for an RSI breakthrough seems to be accelerating, which could make delaying IPO more advantageous." This is because "technological and world changes might happen in astonishing ways, and there might be good reasons to stay private during that time." Last week, Anthropic released a blog post warning about the potential dangers if their AI model Claude achieves full recursive self-improvement and begins to build more powerful descendants, which could lead to humans losing control of AI systems. Reports citing sources also mentioned that according to an internal message from OpenAI's Chief Scientist, Ilya Sutskever, OpenAI is preparing to release a new model codenamed 5.6, which is expected to have "significant improvements" compared to GPT-5.5.