AI Sec Watch

This academic paper explores how Software Bill of Materials (SBOMs, detailed lists of all software components used in a project) can be extended to cover agentic AI systems (AI systems that can independently make decisions and take actions). The paper discusses schema extensions, how to organize and orchestrate these agentic components, and methods to evaluate whether AI systems produce reproducible results.

Microsoft and OpenAI amended their partnership agreement to clarify their long-term relationship and how they will work together on AI development. Key changes include OpenAI gaining freedom to sell products through any cloud provider (not just Microsoft's Azure), Microsoft receiving a non-exclusive license to OpenAI's technology through 2032, and changes to how the companies share revenue. The amendment aims to give both companies flexibility while maintaining their collaborative work on building large-scale AI systems.

Choco, an AI-powered food distribution platform serving over 100,000 buyers, uses OpenAI APIs to power AI agents that automate order processing from multiple input types (emails, texts, images, voice calls). OrderAgent and VoiceAgent convert unstructured customer inputs into structured ERP (enterprise resource planning, a system that manages business operations) orders by learning from each customer's ordering history, achieving up to 50% reduction in manual work and error rates below 1-5%.

A vulnerability (CVE-2026-7061) was found in Toowiredd chatgpt-mcp-server version 0.1.0 that allows OS command injection (running unauthorized system commands on a server through malicious input) in the MCP/HTTP component. The flaw can be exploited remotely by attackers, and public exploit code is already available, but the developers have not yet responded to the security report.

This research paper evaluates whether multiple AI agents working together can effectively help identify privacy threats in software systems using LINDDUN GO, a structured methodology for privacy threat modeling (a process of identifying ways a system could leak or misuse personal data). The study, published in July 2026, examines whether collaborative multi-agent LLM (large language model) systems can improve the quality and completeness of privacy threat identification compared to single AI agents or human analysis.

Elon Musk is suing Sam Altman and OpenAI in court, claiming that Altman broke the company's original founding agreement. The lawsuit focuses on OpenAI's early years when it was started as a nonprofit, and the trial could influence the direction of AI development in the tech industry.

A security flaw called CVE-2026-7020 was found in Ollama versions up to 0.20.2 that allows path traversal (an attack where someone manipulates file paths to access files they shouldn't be able to reach) through the digestToPath function in the Tensor Model Transfer Handler component. An attacker can exploit this remotely, though it requires high complexity to perform, and the vulnerability details have been released publicly.

n8n-mcp (a tool for connecting AI systems to external services) was logging sensitive information like passwords and API keys when running in HTTP mode (a way to communicate over the internet). When authenticated users made requests to call tools, their secret credentials were written to server logs before being hidden, which could expose them if logs were shared or accessed by unauthorized people. The issue only affected HTTP mode and required authentication, so it couldn't be exploited by random internet users.

LiteLLM had a security flaw in two test endpoints (`POST /mcp-rest/test/connection` and `POST /mcp-rest/test/tools/list`) that allowed authenticated users to run arbitrary commands on the server. These endpoints accepted server configurations including command and arguments, and would execute them as subprocesses with the proxy's privileges, even for users with low-level permissions.