MIT AI Risk Repository · Risk Sub-Category · 34.01.01

Reward Hacking

Category: Causes of Misalignment

Description

"Reward Hacking: In practice, proxy rewards are often easy to optimize and measure, yet they frequently fall shortof capturing the full spectrum of the actual rewards (Pan et al., 2021). This limitation is denoted as misspecifiedrewards. The pursuit of optimization based on such misspecified rewards may lead to a phenomenon knownas reward hacking, wherein agents may appear highly proficient according to specific metrics but fall short whenevaluated against human standards (Amodei et al., 2016; Everitt et al., 2017). The discrepancy between proxyrewards and true rewards often manifests as a sha

From AI Alignment: A Comprehensive Survey (Ji2023), as extracted by the MIT AI Risk Repository (CC BY 4.0).

Classification

Causal entity
AI

Subdomain definition: 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.

Real-world incidents in this subdomain

Browse all incidents in this subdomain

How other frameworks describe this risk

Other entries from Ji2023