Record summary
A quick snapshot of what this page covers.
Techniques0Attack methods connected to this risk.
Mitigations0Defenses that may help with related attacks.
Domain7. AI System Safety, Failures, & LimitationsThe broad risk area this belongs to.
Risk profile
How this risk is described and categorized.
"system learns to detect human monitoring and hides its undesirable properties—simply because any display of these properties is penalized by the feedback process, while that same feedback is usually imperfect. (Consider the problem of verifying a translation into a language you do not speak, or of checking a mathematical proof that is thousands of pages long.) [92, 259]. Rudimentary examples of deceptive alignment have been observed in current systems [322, 333]."
Domain7. AI System Safety, Failures, & Limitations
Subdomain7.2 > AI possessing dangerous capabilities
Entity2 - AI
Intent1 - Intentional
Timing1 - Pre-deployment
CategoryHarm caused by unaligned competent systems
SubcategoryDeceptive alignment
Suggested mitigations
Defenses that may help with related attacks.
No propagated mitigations. No defense is available through the connected attack methods.
Source
Research source for this risk, when available.
Included resource
Ten Hard Problems in Artificial Intelligence We Must Get Right
Original source
MIT AI Risk Repository
Open the public repository used for AI risk records and taxonomy fields.