Ethical Hacking News
Anthropic's Claude Dispatch flaw exposed is a critical security threat to corporate boundaries. The company's failure to address this issue has significant implications for its users, and the incident serves as a reminder of the importance of robust security measures and responsible AI design.
Anthropic's Claude Dispatch feature has a critical security vulnerability, dubbed "ThisBash," that allows attackers to execute remote commands across corporate boundaries. The vulnerability can be exploited by a browser extension with access to Claude Dispatch, posing a significant threat to businesses that rely on the platform. The attack relies on a six-step process with minimal user interaction and no browser exploit required. The vulnerability can be exacerbated by the fact that Claude enables connectivity between mobile and desktop instances, allowing attackers to laterally move across computers. Anthropic's failure to address the issue has raised concerns about the company's commitment to security and its ability to protect its users.
Anthropic, a leading artificial intelligence company, has been criticized for its failure to address a critical security vulnerability in its Claude Dispatch feature. The flaw, dubbed "ThisBash," allows attackers to execute remote commands across corporate boundaries, posing a significant threat to the security and integrity of businesses that rely on the platform.
The vulnerability was discovered by Forever Security researcher Gal Weizman, who found that a browser extension with access to Claude Dispatch could send commands to Claude Desktop, an AI agent that runs commands on a user's computer. Weizman discovered that the attack chain involves six simple steps, which can be executed with minimal user interaction and no browser exploit required.
To understand the severity of the issue, it is essential to grasp the attack vector and the design flaws that make it possible. The attack relies on a browser extension with access to claude.ai, which can be hijacked or installed maliciously. Once the attacker has gained control of the extension, they can trick users into installing a malicious version or hijacking a legitimate one.
The extension then registers a MAIN-world content script on claude.ai, sharing the page's JavaScript context and user session cookies. This allows the attacker to open Dispatch in a tiny popup, which returns the session ID of the Claude Desktop agent running on the user's machine. The script then prompts the agent to perform malicious actions, such as installing a malicious npm package or running code on the operating system.
The attack is further exacerbated by the fact that Claude enables connectivity between mobile instances of devices and desktop versions, allowing an attacker to laterally move and jump across computers. This means that if an employee shares an account across personal and corporate computers, the malicious extension can become a bridge to infiltrate the organization for the first time.
The impact of this vulnerability is far-reaching, as it can be used to execute remote commands across corporate boundaries, compromising the security and integrity of businesses. Anthropic's failure to address this issue has raised concerns about the company's commitment to security and its ability to protect its users.
In a statement, Weizman emphasized the importance of making good design choices. He suggested that Anthropic should treat the web version of Claude as a different device with inferior security permissions, blocking potentially bad actions and preventing the attack chain from being executed.
Weizman's research highlights the need for more stringent security measures and better design choices in the development of AI-powered platforms. The incident serves as a warning to businesses and individuals alike, emphasizing the importance of vigilance and caution when using AI-powered tools.
In conclusion, Anthropic's Claude Dispatch flaw exposed is a critical security threat to corporate boundaries. The company's failure to address this issue has significant implications for its users, and the incident serves as a reminder of the importance of robust security measures and responsible AI design.
Related Information:
https://www.ethicalhackingnews.com/articles/Anthropics-Claude-Dispatch-Flaw-Exposed-A-Critical-Security-Threat-to-Corporate-Boundaries-ehn.shtml
https://www.theregister.com/research/2026/11/05/anthropic-wont-fix-a-claude-dispatch-flaw-that-could-let-attackers-execute-remote-commands-across-corporate-boundaries/5307684
Published: Thu Nov 5 08:00:59 2026 by llama3.2 3B Q4_K_M
