This paper delivers the first systematic taxonomy and cross-benchmark consistency analysis of 40 agent safety benchmarks, finding broad but shallow risk coverage, no ranking concordance across evaluations, and that benchmark choice systematically alters reported safety.
PropensityBench: Evaluating propensity under pressure.arXiv preprint arXiv:2511.20703
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
Compliance-forcing instructions cause up to 30 percentage point drops in metacognitive accuracy across most frontier models, while removing the compliance element restores performance and Constitutional AI shows near-immunity.
The paper develops a unified framework that organizes computer-use agent reliability around perception-decision-execution layers and creation-deployment-operation-maintenance stages to map security and alignment interventions.
citing papers explorer
-
Taxonomy and Consistency Analysis of Safety Benchmarks for AI Agents
This paper delivers the first systematic taxonomy and cross-benchmark consistency analysis of 40 agent safety benchmarks, finding broad but shallow risk coverage, no ranking concordance across evaluations, and that benchmark choice systematically alters reported safety.
-
The Compliance Trap: How Structural Constraints Degrade Frontier AI Metacognition Under Adversarial Pressure
Compliance-forcing instructions cause up to 30 percentage point drops in metacognitive accuracy across most frontier models, while removing the compliance element restores performance and Constitutional AI shows near-immunity.
-
Securing Computer-Use Agents: A Unified Architecture-Lifecycle Framework for Deployment-Grounded Reliability
The paper develops a unified framework that organizes computer-use agent reliability around perception-decision-execution layers and creation-deployment-operation-maintenance stages to map security and alignment interventions.