OpenAI is trying to pull off two hard things at once: go public at an eye-watering valuation, and convince regulators it can keep its systems secure.
According to a report published by The Week (via Bing News), the company confidentially filed for a US initial public offering in June 2026 and is aiming to list by 2027 at a potential valuation of $1 trillion. Confidential filings are common — they let a company start the regulatory process without showing its financials to competitors until closer to the listing date.
That timeline is now running alongside a security problem. Yahoo, citing coverage of the case, reports that Alabama's attorney general has subpoenaed OpenAI over a hacking incident. A subpoena is a legal demand for documents and information; it is not an accusation of wrongdoing, but it does mean a state law enforcement office wants a closer look.
Separately, OpenAI has been acting against misuse of its own products. The Caspian Post reports that the company blocked Russia-linked ChatGPT accounts. The Register covered the same enforcement action in characteristically blunt terms, framing it as OpenAI banning Russian users for using its tools to generate low-quality mass content.
The common thread is trust. An IPO prospectus has to disclose material risks, and security incidents plus active state investigations are exactly the kind of disclosure that invites questions from regulators and investors alike. The open question raised by The Week's headline — whether the probe delays the offering — has no public answer yet.
Why it matters: if OpenAI lists near $1 trillion, ordinary investors and pension funds will own a piece of the company shaping everyday AI, and how it handles security scrutiny now sets the terms of that bet.