SB 1047 debate turns on whether rogue-AI risk or catastrophic misuse came first

deanwball · x · 2026-07-23

Dean Ball argues that the existence of mechanisms like a Frontier Model Division and kill switches in SB 1047 shows that loss of control / rogue AI was a real motivation behind the bill.

He adds two nuances:

The bill was initially more driven by rogue-AI concerns, but some of that was later dialed back.
In his view, today’s models look less likely to “go rogue” than they did in 2024.

The exchange is about how to read the bill’s intent: whether catastrophic misuse or loss-of-control risks were the primary driver.

Original post →

SB 1047 debate turns on whether rogue-AI risk or catastrophic misuse came first

More from AGI Musings