Bannon and Sanders, same stage

A right-wing provocateur and a progressive senator are teaming up over AI risk, joined by religious leaders, lawmakers, and Joseph Gordon-Levitt.

A resignation lit the fuse

An Anthropic researcher quit publicly last week and said the people building these systems think the technology could kill everyone within the decade.

“earnestly believe it could kill us all by the end of the decade”

Not a fringe worry anymore

Over the weekend, leading AI researchers and executives said the technology is moving dangerously fast and asked for measures to rein it in.

“It sounds like science fiction. It is the stuff of science fiction. But it's not a fringe position.”

Trump is a hard no

Without the president, substantive AI rules have no path. He called the pushback a conspiracy on Truth Social and said only China benefits from it.

“There is a SICK conspiracy going on against AI and Data Centers, and the only one happy about it is China.”

Option one: ban it

Federal law could bar development of systems that exceed human performance, or pause that work until a new agency exists to watch over it.

Option two: real auditors

A government agency, nonprofit, or industry body could review models before public release. Today's version is voluntary, secretive, and staffed by unclear expertise.

Option three: make them pay

Congress could rewrite liability law so companies own the harm their systems and agents cause. A bankrupting lawsuit is a strong reason to slow down.

The clock, not the law

All three options already have precedent in federal regulation. The obstacle is agreement, and one researcher fears the window closes while Congress argues.

“Washington, DC, moves far, far slower than Silicon Valley.”