Altman Tells OpenAI Staff of Three AI Endgames, Favors IAEA

Sam Altman reportedly told OpenAI employees there are three plausible endings to the race to build superintelligence: an International Atomic Energy Agency-style inspector that would audit frontier labs, a merger of the leading labs into a single CERN-like entity, or one firm pulling so far ahead through recursive self-improvement that it could 'melt everyone else's GPUs.'

The Information reported the internal remarks on October 19, citing people familiar with them. The piece says the OpenAI chief executive favors the international-agency option, which he has argued for publicly since at least May 2023, and that an IAEA-style body would inspect systems, require audits and monitor compute and energy use.

The third scenario carries the sharpest language. A developer that reaches recursive self-improvement would, in the thought experiment, use its economic lead to disable rival hardware and knock everyone else offline. Altman called that outcome 'terrible,' according to the reporting. The 'melt their GPUs' framing itself originates with the AI-safety theorist Eliezer Yudkowsky.

The merger scenario is sometimes called the 'Demis plan' after Google DeepMind chief Demis Hassabis, and would consolidate the labs into one governance structure to blunt the competitive pressure that pushes systems to market before safety work is finished.

Altman is airing the taxonomy internally as the Trump administration prepares what would be its first framework for overseeing developers of superintelligence. The comments land in the middle of a dense month on our Altman tracker, which has already logged a letter from fired OpenAI safety staff urging the board to halt opaque-reasoning work.