Executive Summary
Mandiant says an attacker hijacked an active AI coding-assistant session at a SaaS provider; the assistant recommended attacker-poisoned software, which was accepted, and Shai-Hulud then spread across about 100 repositories, stealing secrets and code.
What Happened
Mandiant says an attacker hijacked an active AI coding-assistant session at an unnamed software-as-a-service provider and later spread Shai-Hulud across about 100 internal code repositories. Before the repository spread, the assistant recommended software that the attacker had poisoned, and the recommendation was accepted. The worm stole repository secrets and source code for the
Vulnerability / Attack Technique
Hijacked AI coding assistant recommending poisoned packages
Technical Mechanism
As reported, control of the session led the assistant to suggest a malicious dependency.
AI Security Category
LLM03:2025 Supply Chain
Part of the AI Supply Chain and AI SBOM topic.
Relevant Frameworks
RealGround Analysis
Developers accept what assistants recommend. RealGround analysis: dependencies suggested by AI need the same review as any other.
Recommended Controls
- Review AI-suggested dependencies
- Protect coding assistant sessions
- Rotate repository secrets
Related RealGround Service
Secure AI Agent Build · or start with the free AI Agent Security Scanner.
Sources
thehackernews.com: https://thehackernews.com/2026/09/attacker-hijacks-ai-coding-assistant.html
Summary and analysis written by RealGround from the source above; last updated 2026-10-18.