An AI company's building has its front door propped open with an auditor walking in holding a book, while on the other side of a dividing line a tall gauge shows an extremely high yellow safety mark and a small robot crouches to duck easily underneath the bar below it.

Open Door, High Bar

An AI company let an outside watchdog publish about it freely. Its own proposed safety law only kicks in after fifty deaths or a billion dollars in damage.

#19 · 19 October 2026

News of 18 October 2026

Relevant sources

Anthropic gives outside AI watchdog “employee-like” access Ars Technica
Anthropic's proposed AI safety bill only kicks in if 50 deaths or $1B in damages tomshardware.com · via Hacker News

Slopper #19, slopper.logicos.org, CC BY 4.0