OpenAI, according to the New York Times, is considering postponing its IPO until next year. The logic behind it is fairly clear to me: the company is targeting a valuation of up to one trillion dollars and CFO Sarah Friar is aiming for a 2027 listing. Advisors reportedly gave management a choice — either wait until 2027 and go public at the trillion valuation, or lower the target valuation and move faster. Sam Altman responded that any change to that trillion figure is unacceptable. That says a lot about the company: it doesn’t feel pressured or cash‑strapped, it retains negotiating power and can afford to wait for the ideal window. From a position of strength, there’s no rush.

Second, and more interesting to me, is the paragraph about the government. The Trump administration asked OpenAI to phase the rollout of the new GPT 5.6 model over safety concerns — the model is initially being provided only to selected partners in a limited preview, and the government is reportedly approving access on a customer‑by‑customer basis. This is a much more important signal than an IPO date. It means the most advanced AI models are beginning to be treated as strategic technology subject to state control, similar to chips or weapons. That changes the game for the whole sector — regulation ceases to be theoretical and becomes a real operational factor.