pith. sign in

When actions go off-task: Detecting and correcting misaligned actions in computer-use agents

2 Pith papers cite this work. Polarity classification is still indexing.

2 Pith papers citing it
abstract

Computer-use agents (CUAs) have made tremendous progress in the past year, yet they still frequently produce misaligned actions that deviate from the user's original intent. Such misaligned actions may arise from external attacks (e.g., indirect prompt injection) or from internal limitations (e.g., erroneous reasoning). They not only expose CUAs to safety risks, but also degrade task efficiency and reliability. This work makes the first effort to define and study misaligned action detection in CUAs, with comprehensive coverage of both externally induced and internally arising misaligned actions. We further identify three common categories in real-world CUA deployment and construct MisActBench, a benchmark of realistic trajectories with human-annotated, action-level alignment labels. Moreover, we propose DeAction, a practical and universal guardrail that detects misaligned actions before execution and iteratively corrects them through structured feedback. DeAction outperforms all existing baselines across offline and online evaluations with moderate latency overhead: (1) On MisActBench, it outperforms baselines by over 15% absolute in F1 score; (2) In online evaluation, it reduces attack success rate by over 90% under adversarial settings while preserving or even improving task success rate in benign environments.

citation-role summary

background 1

citation-polarity summary

fields

cs.CL 1 cs.CR 1

years

2026 2

verdicts

UNVERDICTED 2

roles

background 1

polarities

support 1

representative citing papers

Triaging Threats to Specialized Guardrails

cs.CR · 2026-05-29 · unverdicted · novelty 6.0

Introduces GuardZoo benchmark and RouteGuard router-expert system showing monolithic guardrails suffer task interference while specialized routing improves threat detection and generalization.

citing papers explorer

Showing 2 of 2 citing papers.