MIT AI Risk Repository · Additional evidence · 68.03.00c

Sudden loss of control

Open the MIT Risk Navigator All entries from Chin2025

Description

"The key characteristic of this risk is that a single AI agent competently takes actions that lead to a catastrophic outcome. It does not require the AI to be intentional in its actions, only competent enough to make and execute plans that ultimately result in a catastrophe."

From Dimensional Characterization and Pathway Modeling for Catastrophic AI Risks (Chin2025), as extracted by the MIT AI Risk Repository (CC BY 4.0).

Classification

Domain: —
Subdomain: —
Causal entity: —
Intent: —
Timing: —

Other entries from Chin2025

CBRN Risk Category · 4.2
CBRN Additional evidence
CBRN Additional evidence
CBRN Additional evidence
Cyber offense Risk Category · 4.2
Cyber offense Additional evidence
Cyber offense Additional evidence
Cyber offense Additional evidence
Sudden loss of control Risk Category · 7.1
Sudden loss of control Additional evidence
Sudden loss of control Additional evidence
Sudden loss of control Additional evidence