AI Sec Watch

The European Commission is recruiting up to 60 independent experts for a scientific panel to advise on general-purpose AI (GPAI, large AI models designed for many tasks) under the EU AI Act. The panel will assess systemic risks (widespread dangers affecting multiple countries or many users), classify AI models, and issue alerts when AI systems pose significant dangers to Europe. Applicants need a PhD in a relevant field, proven AI research experience, and independence from AI companies, with the deadline set for September 14th.

This article compares traditional application security (AppSec) practices with AI security, noting that familiar principles like input validation and authentication apply to both, but AI systems introduce unique risks. New attack types specific to AI, such as prompt injection (tricking an AI by hiding instructions in its input), model poisoning (tampering with training data), and membership inference attacks (determining if specific data was in training), require security engineers to develop new defensive strategies beyond traditional code-level vulnerability management.

Cursor, a code editor designed for AI-assisted programming, had a security flaw in versions before 0.51.0 where JSON files could automatically trigger web requests without user approval. An attacker could exploit this, especially after a prompt injection attack (tricking the AI with hidden instructions in its input), to make the AI agent send data to a malicious website.

CVE-2025-32711 is a command injection vulnerability (a weakness where an attacker tricks a program into running unintended commands) in Microsoft 365 Copilot that allows an unauthorized attacker to disclose information over a network. The vulnerability has a CVSS severity score of 4.0 (a moderate rating on a 0-10 scale where 10 is most severe). Microsoft has published information about this vulnerability, but the provided source does not contain specific technical details about the attack or its impact.

FastGPT is an open-source platform for building AI workflows and chatbots that uses a sandbox (an isolated container designed to safely run untrusted code). Versions before 4.9.11 had weak isolation that allowed attackers to escape the sandbox by using overly permissive syscalls (system calls, which are requests programs make to the operating system), letting them read files, modify files, and bypass security restrictions. The vulnerability is fixed in version 4.9.11 by limiting which system calls are allowed to a safer set.

Generative AI (AI systems that create new text, code, or images) is a double-edged sword in cybersecurity, helping both defenders and attackers. The case study of a fictional insurance company shows how GenAI can be used to launch cyberattacks (malicious attempts to breach computer systems) and also to defend against them, creating a difficult choice for IT leaders about whether to use AI as a defensive tool or risk falling behind attackers who already have it.

As AI development has grown rapidly, organizations struggle with how to actually put responsible AI practices into action beyond just making promises about it. This article describes how two organizations created a five-phase process to embed responsibility pledges (formal commitments to use AI ethically) into their daily practices using a systems approach (treating responsibility as interconnected parts of the whole organization rather than isolated efforts).

The mcp-com-server is a tool that connects the Model Context Protocol (MCP, a standard for AI systems to interact with external tools) to COM (Component Object Model, Microsoft's decades-old system for sharing functionality across programs on Windows). This allows an AI like Claude to automate Windows and Office tasks, such as creating Excel files and sending emails, by dynamically discovering and controlling COM objects. The main security risk is that COM can access dangerous operations like file system access, so the server uses an allowlist (a list of approved COM objects that are permitted to run) to restrict which COM objects can be instantiated.

Skyvern through version 0.1.85 has a vulnerability where attackers can inject malicious code into the Prompt field of workflow blocks through SSTI (server-side template injection, where untrusted input is processed as code by the server's template engine). Authenticated users can craft special expressions in Jinja2 templates (a template system that evaluates code on the server) that aren't properly cleaned up, allowing them to execute commands on the server without direct feedback, a capability known as blind RCE (remote code execution).