DeepMind researcher warns of declining CoT monitorability, a core misalignment safety tool
tomekkorbak · x · 2026-09-04
Researcher tomekkorbak says he is deeply worried by the trend of decreasing chain-of-thought (CoT) monitorability: CoT monitoring is a core part of the misalignment safety strategy with no good substitute today.
Follow-up findings: CoT controllability increases over the course of RL training (unlike previous models) and is strongly correlated with no-CoT capabilities across several model generations. The team will keep tracking why monitorability is declining, try to reverse the trend, and soon share thinking on the minimum level of monitorability required.
More from Safety
- GitHub HQ Hosts Open Source AI + Policy Talk with Mozilla Co-founder Mitchell Baker — kevinsxu · 2026-10-21
- Researcher Finds Two Windows 0-Days, First in 76 Minutes with a Frontier Model — moyix · 2026-10-21
- Stanford's Noah Goodman shares interactive AI safety concept map — stanfordnlp · 2026-10-21
- AI agents are overwhelming Gmail's spam filter, researcher warns — LuizaJarovsky · 2026-10-21
- AI safety researcher Sören Mindermann moves to Tsinghua, co-founds Beijing AI Safety Hub — dhadfieldmenell · 2026-10-21
- Talos researcher: AI agent attacks today are loud pentests — quiet swarms are the real threat — matthew_d_green · 2026-10-21