AI & LLM Vulnerabilities

OpenClaw had a vulnerability where its hook endpoint (`POST /hooks/agent`) accepted session keys (identifiers for conversation contexts) directly from user requests, allowing someone with a valid hook token to inject messages into any session they could guess or derive. This could poison conversations with malicious prompts that persist across multiple turns. The vulnerability affected versions 2.0.0-beta3 through 2026.2.11.

Cursor, a code editor designed for programming with AI, had a sandbox escape vulnerability in versions before 2.5 where a malicious agent (an attacker using prompt injection, which is tricking an AI by hiding instructions in its input) could write to unprotected .git configuration files, including git hooks (scripts that run automatically when Git performs certain actions). This could lead to RCE (remote code execution, where an attacker runs commands on a system they don't control) when those hooks were triggered, with no user action needed.

A reflected XSS vulnerability (a type of attack where malicious code is injected into a website and executed in a user's browser) was found in the AI Playground's OAuth callback handler (the code that processes login responses). The vulnerability allowed attackers to craft malicious links that, when clicked, could steal a user's chat history and access connected MCP servers (external services integrated with the AI system) on the victim's behalf.

CVE-2026-1669 is a vulnerability in Keras (a machine learning library) versions 3.0.0 through 3.13.1 that allows attackers to read arbitrary files on a system by uploading a specially crafted model file that exploits HDF5 external dataset references (a feature of HDF5, a file format commonly used to store large amounts of numerical data). An attacker could use this to access sensitive information stored on the affected computer.

sf-mcp-server, a tool that connects Salesforce to Claude for Desktop, has a command injection vulnerability (CWE-78, a flaw where attackers inject malicious commands into user input). The vulnerability exists because the software unsafely uses child_process.exec (a function that runs shell commands) with user-controlled input, allowing attackers to execute arbitrary shell commands with the server's privileges.

CVE-2026-21523 is a time-of-check time-of-use (TOCTOU) race condition (a vulnerability where an attacker exploits the gap between when a system checks permissions and when it uses a resource) in GitHub Copilot and Visual Studio that allows an authorized attacker to execute code over a network. The vulnerability has not yet received a CVSS severity rating from NIST.

GitHub Copilot contains a command injection vulnerability (CVE-2026-21516), which is a flaw where special characters in user input are not properly filtered, allowing an attacker to execute code remotely on a system. The vulnerability was reported by Microsoft Corporation and has a CVSS score pending assessment.

CVE-2026-21257 is a command injection vulnerability (a flaw where attackers can insert malicious commands into an application) found in GitHub Copilot and Visual Studio that allows an authorized attacker to gain elevated privileges over a network. The vulnerability stems from improper handling of special characters in commands. As of the source date, a CVSS severity score (a 0-10 rating of how severe a vulnerability is) had not yet been assigned by NIST.