Agency

Category

Risk Domain

7AI System Safety, Failures & Limitations

AI systems that develop, access, or are provided with capabilities that increase their potential to cause mass harm through deception, weapons development and acquisition, persuasion and manipulation, political strategy, cyber-offense, AI development, situational awareness, and self-proliferation. These capabilities may cause mass harm due to malicious human actors, misaligned AI systems, or failure in the AI system.

"This section catalogs the risk sources and risk management measures related to agentic AI systems. We categorize these into the following groups: goal- directedness, deception, situational awareness, self-proliferation, and persuasion"(p. 28)

Entity— Who or what caused the harm

Human

Due to a decision or action made by humans

AI system

Due to a decision or action made by an AI system

Other

Due to some other reason or is ambiguous

Not coded

Intent— Whether the harm was intentional or accidental

Intentional

Due to an expected outcome from pursuing a goal

Unintentional

Due to an unexpected outcome from pursuing a goal

Other

Without clearly specifying the intentionality

Not coded

Timing— Whether the risk is pre- or post-deployment

Pre-deployment

Occurring before the AI is deployed

Post-deployment

Occurring after the AI model has been trained and deployed

Other

Without a clearly specified time of occurrence

Not coded

Supporting Evidence (1)

"These risk items are related to behaviors associated with agentic systems using a GPAI as a base model or other component that is tasked with achieving an objective by manipulate its environment. They should not be confused with autonomy, where an AI system is free to take actions without full human supervision."