DeepMind researcher says AI misuse and system-level risks matter more than scheming models

DeepMind's Seb Krier argues that AI misuse by non-state actors and institutional failure pose greater risks than autonomous AI scheming, and endorses the view that complex multi-agent systems demand safety work beyond model-level alignment.

2026-11-07 ~ 2026-11-07 · 2 related posts

• Seb Krier lays out his AI safety views: misuse and institutional failure over scheming AIs — sebkrier · 2026-11-07
• Safety researcher: complex agentic systems make model-scoped alignment look woefully insufficient — sebkrier · 2026-11-07