Abstract:Computer-use agents (CUAs) promise to automate complex tasks across operating systems (OS) and the web, but remain vulnerable to indirect prompt injection. Current evaluations of this threat either lack support realistic but controlled environments or ignore hybrid web-OS attack scenarios involving both interfaces. To address this, we propose RedTeamCUA, an adversarial testing framework featuring a novel hybrid sandbox that integrates a VM-based OS environment with Docker-based web platforms. Our sandbox supports key features tailored for red teaming, such as flexible adversarial scenario configuration, and a setting that decouples adversarial evaluation from navigational limitations of CUAs by initializing tests directly at the point of an adversarial injection. Using RedTeamCUA, we develop RTC-Bench, a comprehensive benchmark with 864 examples that investigate realistic, hybrid web-OS attack scenarios and fundamental security vulnerabilities. Benchmarking current frontier CUAs identifies significant vulnerabilities: Claude 3.7 Sonnet | CUA demonstrates an ASR of 42.9%, while Operator, the most secure CUA evaluated, still exhibits an ASR of 7.6%. Notably, CUAs often attempt to execute adversarial tasks with an Attempt Rate as high as 92.5%, although failing to complete them due to capability limitations. Nevertheless, we observe concerning high ASRs in realistic end-to-end settings, with the strongest-to-date Claude 4.5 Sonnet | CUA exhibiting the highest ASR of 60%, indicating that CUA threats can already result in tangible risks to users and computer systems. Overall, RedTeamCUA provides an essential framework for advancing realistic, controlled, and systematic analysis of CUA vulnerabilities, highlighting the urgent need for robust defenses to indirect prompt injection prior to real-world deployment.
Submission history
From: Zeyi Liao [view email]
[v1]
Wed, 28 May 2025 03:42:09 UTC (3,055 KB)
[v2]
Sun, 1 Jun 2025 00:04:07 UTC (3,003 KB)
[v3]
Wed, 15 Oct 2025 04:01:38 UTC (3,026 KB)
[v4]
Mon, 16 Feb 2026 14:55:16 UTC (3,036 KB)
[v5]
Sun, 1 Mar 2026 07:50:09 UTC (3,061 KB)