Open access
Aug 2026
Understanding Cognition-Induced Risks in Agentic AI Systems
This work systematically analyzes risks induced by expanding cognitive capabilities, following a three-level framework defined by their cognitive scope, from physical cognition to social cognition, and finally to self-referential cognition.
Guanchu Wang, Qinuo Li, Mengnan Du et al.
· IEEE Intelligent Systems · 0 citations