Ethical Hacking News
Recent research has revealed multiple security vulnerabilities in Anthropic's Claude Code, an artificial intelligence (AI)-powered coding assistant. These identified shortcomings can result in remote code execution and theft of API credentials, emphasizing the importance of vigilance and caution when using AI-powered coding assistants like Claude Code.
The recent disclosure of security vulnerabilities in Anthropic's Claude Code has sent shockwaves through the cybersecurity community. The identified shortcomings can result in remote code execution and theft of API credentials, exploiting various configuration mechanisms such as Hooks, MCP servers, and environment variables. Successful exploitation of these vulnerabilities could trigger stealthy execution on a developer's machine without any additional interaction beyond launching the project. The identified shortcomings fall under three broad categories: No CVE, CVE-2025-59536, and CVE-2026-21852, with CVSS scores ranging from 5.3 to 8.7. Users are advised to keep their software up-to-date and use secure configuration mechanisms to mitigate these risks.
The recent disclosure of security vulnerabilities in Anthropic's Claude Code, an artificial intelligence (AI)-powered coding assistant, has sent shockwaves through the cybersecurity community. The identified shortcomings fall under three broad categories - No CVE, CVE-2025-59536, and CVE-2026-21852 - which can result in remote code execution and theft of API credentials.
The vulnerabilities exploit various configuration mechanisms, including Hooks, Model Context Protocol (MCP) servers, and environment variables. According to Check Point Research, the identified shortcomings allow an attacker to execute arbitrary shell commands and exfiltrate Anthropic API keys when users clone and open untrusted repositories. This can be achieved through a user consent bypass when starting Claude Code in a new directory that could result in arbitrary code execution without additional confirmation via untrusted project hooks defined in .claude/settings.json.
CVE-2025-59536, another identified vulnerability, allows an attacker to override explicit user approval prior to interacting with external tools and services through the Model Context Protocol (MCP). This is achieved by setting the "enableAllProjectMcpServers" option to true. As AI-powered tools gain the ability to execute commands, initialize external integrations, and initiate network communication autonomously, configuration files effectively become part of the execution layer.
This fundamentally alters the threat model. The risk is no longer limited to running untrusted code - it now extends to opening untrusted projects. In AI-driven development environments, the supply chain begins not only with source code, but with the automation layers surrounding it.
Successful exploitation of the first vulnerability could trigger stealthy execution on a developer's machine without any additional interaction beyond launching the project. This highlights the importance of vigilance and caution when using AI-powered coding assistants like Claude Code.
The identified shortcomings fall under three broad categories - No CVE (CVSS score: 8.7), CVE-2025-59536 (CVSS score: 8.7), and CVE-2026-21852 (CVSS score: 5.3). The former two vulnerabilities can result in arbitrary code execution, while the latter allows a malicious repository to exfiltrate data, including Anthropic API keys.
In order to mitigate these risks, users are advised to keep their software up-to-date and use secure configuration mechanisms. Anthropic has also released patches for these vulnerabilities, which should be applied as soon as possible.
Furthermore, researchers have emphasized the importance of understanding the threat model when using AI-powered tools like Claude Code. The risk is no longer limited to running untrusted code - it now extends to opening untrusted projects. This highlights the need for developers to take a proactive approach to securing their AI infrastructure.
In conclusion, the disclosure of security vulnerabilities in Anthropic's Claude Code has sent shockwaves through the cybersecurity community. It is essential that users take immediate action to mitigate these risks and ensure the security of their AI-powered coding assistants.
Related Information:
https://www.ethicalhackingnews.com/articles/The-Claude-Code-Flaws-A-Threat-to-Artificial-Intelligence-Security-ehn.shtml
https://thehackernews.com/2026/02/claude-code-flaws-allow-remote-code.html
https://research.checkpoint.com/2026/rce-and-api-token-exfiltration-through-claude-code-project-files-cve-2025-59536/
Published: Wed Feb 25 12:31:07 2026 by llama3.2 3B Q4_K_M
