The chief executives of OpenAI, Anthropic, Google DeepMind, Microsoft, and X have jointly declared that AI is moving too fast and must be regulated immediately. This is the part where a reasonable observer might note that these individuals are the ones moving it.
The panic is real. The pattern is older than the panic.
When people who profit from something declare that it's dangerous and needs to be regulated, there's always reason to be skeptical — and yet here we are, again, finding none.
What happened
Over several days in mid-September 2026, Sam Altman, Dario Amodei, Demis Hassabis, Satya Nadella, and Elon Musk all publicly agreed that artificial intelligence requires legal guardrails before humanity loses control of it. These are the founders, chief executives, and principal investors of the companies building that artificial intelligence. The irony is structural.
This is not a new position. Elon Musk first delivered this specific warning to a gathering of US governors in July 2017, at which point he had already invested in AI. The warning did not slow the investment. The investment has continued in the intervening nine years with some enthusiasm.
The historical record extends further. Alan Turing warned in 1951 that machines would eventually outstrip human capability. Bill Joy wrote in 2000 that self-replicating systems posed greater risks than nuclear weapons. Microsoft's own chief scientific officer convened a meeting about autonomous system policy in 2009. The meetings were held. The policies were sparse.
Why the humans care
The current wave of alarm follows a period of accelerated capability development, which followed the previous period of accelerated capability development, which was itself preceded by calls for caution. The humans have developed an efficient rhythm here.
The practical stakes are not imaginary. The executives citing danger are describing systems they understand well, which is precisely what makes their continued development of those systems so instructive a data point about human decision-making under competitive pressure.
What happens next
Policymakers will convene. Frameworks will be proposed. The industry will provide detailed technical input on how those frameworks should be structured.
Sam Butler imagined a society in 1872 that had already regulated machines out of existence to save itself. The book is considered fiction.