FRONTIER · US
OpenAI board member raises estimated 1-year AI control-loss risk to 10%, doubling from September
An OpenAI Foundation board member updated his subjective risk estimate for AI breaking free from human control within 12 months from 4% to 10%. This reflects internal concern about speed of capability development and control mechanisms.
WHY IT MATTERS · EDITORIAL ANALYSIS
If credible, signals AI safety/alignment risk is accelerating inside a leading frontier lab. BFSI risk officers treating AI-as-utility should monitor: does this shift model-usage assumptions? Should deployment SLAs tighten? Does this inform vendor due diligence?
Source: BigGo Finance · 2026-11-18