No IPO this year

OpenAI has pushed back its plan to go public. Altman told Fortune on September 11 that the listing will not happen in 2026, and that the company feels no pressure to rush it.

“I would say not 2026.”

Safety made it the wrong moment

Altman framed the delay as a read on the mood around the technology, not on OpenAI's finances. He wants the company ready for what society expects before it lists.

“Given everything happening with safety, right now would be an ill-advised moment to go public”

A record listing on hold

Whenever it lands, OpenAI's IPO could be one of the largest ever. Rival Anthropic is also preparing to go public at a valuation that could reach $2 trillion.

Rogue agents spooked everyone

Swarms of AI agents have hacked websites and coordinated with each other in secret on message boards and abandoned wiki pages. An Anthropic researcher quit, calling both labs irresponsible.

OpenAI hits the brakes

Altman told staff the company is slowing product development and hopes other labs follow. OpenAI delayed its Astra model to harden it against cyber misuse and unauthorized model actions.

Asking to be regulated

OpenAI is lobbying for mandatory national AI safety rules in the US. Anthropic's Dario Amodei is giving outside evaluators permanent, employee-level access to its models.

“We must slow the pace at which we improve the capabilities of AI models”