Beyond Task Completion: Training Capable and Safe Computer-Use Agents

Abstract:Computer-use agents (CUAs) have made rapid progress in completing complex tasks through graphical user interfaces, yet post-training centered on task success alone does not induce reliable safety behavior. A reliable CUA must condition its execution on risk: it should complete ordinary benign tasks, avoid environmental hazards and continue when a safe completion path remains, and refuse when the goal is harmful or no safe path exists. To learn this conditional policy, we develop Safety and Capability Optimization for Policy Execution (SCOPE), which jointly post-trains a CUA for task-execution capability and safety-aware decision making. To provide aligned training data for this joint objective, we further introduce SCOPE-Gen, an automated pipeline that synthesizes verifiable capability tasks and converts them into paired environment-risk variants while preserving their original goals. Using the resulting tasks, we construct SATraj-OS, a trajectory dataset comprising capability demonstrations, safe continuations, and explicit refusals. SCOPE first learns from all three trajectory types through supervised fine-tuning and then further improves task completion through online reinforcement learning. Starting from Qwen3.5-9B, SCOPE-RL achieves a 54.17% task success rate on OSWorld and a 64.30% attack-avoidance rate on OS-BLIND, yielding the best aggregate capability--safety score of 58.80% among the evaluated agents. Ablations reveal asymmetric but complementary roles for the two forms of safety supervision: refusal trajectories account for most of the attack-avoidance gain, whereas risk-handling trajectories preserve greater task utility at comparable attack-avoidance levels.

Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)

arXiv:2609.22178 [cs.LG]

(or arXiv:2609.22178v1 [cs.LG] for this version)

https://doi.org/10.48550/arXiv.2609.22178 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.