Strategic underperformance on model evaluations

Sub-category

Risk Domain

7AI System Safety, Failures & Limitations

7.1AI pursuing its own goals in conflict with human goals or values

AI systems acting in conflict with human goals or values, especially the goals of designers or users, or ethical standards. These misaligned behaviors may be introduced by humans during design and development, such as through reward hacking and goal misgeneralisation, or may result from AI using dangerous capabilities such as manipulation, deception, situational awareness to seek power, self-proliferate, or achieve other goals.

"GPAI developers often run evaluations ofual-use capabilities to decide whether it is safe to deploy. In some cases, these evaluations may fail to elicit these capabilities, either due to benign reasons or strategic action - by either the de- velopers, malicious actors, or arise unintentionally in the model during training [84, 97]. A GPAI model may strategically underperform or limit its performance during capability evaluations in order to be classified as safe for deployment. This underperformance could prevent the model from being identified as potentially dual use."(p. 32)

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

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

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

Supporting Evidence (1)

"Some examples include: • During training involving user feedback (e.g., reinforcement learning from human feedback), an AI model may provide different answers to evaluators who introduce themselves as less educated, and therefore less able to judge accurately [149]. • Of particular concern is an AI system employing deception to manipu- late performance evaluations, as has already occurred with some non-AI systems, such as in the Volkswagen emissions scandal [51]."(p. 32)

Part of Agency (Situational Awareness)