Evaluating Human-AI Safety: A Framework for Measuring Harmful Capability Uplift
This addresses the problem of inadequate safety assessments for frontier AI models, proposing a human-centered framework for developers, researchers, funders, and regulators, though it is incremental as it builds on prior social science research.
The paper argues that AI safety evaluations should shift from static benchmarks to measuring harmful capability uplift, defined as the marginal increase in a user's ability to cause harm with frontier models beyond conventional tools, and provides methodological guidance for systematic measurement.
Current frontier AI safety evaluations emphasize static benchmarks, third-party annotations, and red-teaming. In this position paper, we argue that AI safety research should focus on human-centered evaluations that measure harmful capability uplift: the marginal increase in a user's ability to cause harm with a frontier model beyond what conventional tools already enable. We frame harmful capability uplift as a core AI safety metric, ground it in prior social science research, and provide concrete methodological guidance for systematic measurement. We conclude with actionable steps for developers, researchers, funders, and regulators to make harmful capability uplift evaluation a standard practice.