Atomicity for Agents: Exposing, Exploiting, and Mitigating TOCTOU Vulnerabilities in Browser-Use Agents

TL;DR

This paper identifies and analyzes TOCTOU vulnerabilities in browser-use agents, demonstrating their prevalence and proposing a lightweight validation method to mitigate these timing-based security risks.

Contribution

It provides the first large-scale empirical study of TOCTOU vulnerabilities in browser agents and introduces a practical validation approach to enhance security.

Findings

TOCTOU vulnerabilities are widespread in browser agents.

Pre-execution validation reduces security risks.

Evaluation of 10 open source agents shows common vulnerabilities.

Abstract

Browser-use agents are widely used for everyday tasks. They enable automated interaction with web pages through structured DOM based interfaces or vision language models operating on page screenshots. However, web pages often change between planning and execution, causing agents to execute actions based on stale assumptions. We view this temporal mismatch as a time of check to time of use (TOCTOU) vulnerability in browser-use agents. Dynamic or adversarial web content can exploit this window to induce unintended actions. We present a large scale empirical study of TOCTOU vulnerabilities in browser-use agents using a benchmark that spans synthesized and real world websites. Using this benchmark, we evaluate 10 popular open source agents and show that TOCTOU vulnerabilities are widespread. We design a lightweight mitigation based on pre-execution validation. It monitors DOM and layout changes during planning and validates the page state immediately before action execution. This approach reduces the risk of insecure execution and mitigates unintended side effects in browser-use agents.