We're still in the 'wildcat banking' era of AI security guarantees, waiting for a blowup

curious_vii · x · 2026-08-28

The author argues AI is still in a wildcat-banking era for security guarantees: companies that can't underwrite security-type guarantees will be dead on arrival for anything this general, and it will unfortunately take a blowup for people to learn. The labs and hyperscalers have pockets deep enough to underwrite literal insurance premia (or something insurance-shaped) to pool risk; upstarts don't — unless they have secret security or risk-mitigation sauce, which they don't.

Original post →

More from AGI Musings

AGI Musings channel →