An AI-safety institute endorses the US superintelligence ban bill
MIRI's first endorsement of a specific federal bill is a position signal only; the bill is still a proposal with unknown prospects.
Original event 2026-09-23
On September 23, MIRI published a post formally endorsing the Ban Artificial Superintelligence Act of 2026 introduced by Sanders and Casar, calling it the first legislation with a real chance of stopping the superintelligence threat.
The post, written by Aaron Scher and endorsed by Bourgon, Soares, and Yudkowsky on behalf of MIRI, notes the bill would pause AI training if models show precursor traits such as automating AI R&D or evading shutdown, and includes a compute threshold adjustable for algorithmic progress. MIRI also flags the bill's lack of AI chip tracking provisions.
Note: the bill is introduced legislation, not law; the endorsement is MIRI's own position statement, and this item has not verified the bill text itself.