AI Red Team Engineering notes, research references, and field notes on AI systems, responsible AI, governance, and compliance.
Global search

Search AI Red Team

Search across the complete knowledge base.

Enter at least two characters to begin.
↑↓ select Enter open Browse the full research archive
Accomplish AI · July 23, 2026

SharedRoot: Escaping the Claude Cowork Sandbox

analysis Agent SecurityAI Red TeamingAI Engineering
Why it matters

Accomplish AI demonstrates SharedRoot, a Claude Cowork local-session escape in which an untrusted task reaches guest root through CVE-2026-46331 and then accesses the Mac host because the entire host filesystem is mounted read-write inside the VM. The durable failure is architectural—unprivileged user namespaces, reachable kernel modules, a permissive seccomp filter, an unhardened root broker, and an over-broad host mount—rather than the single kernel bug; Cowork now defaults to cloud execution.

My takeaway: Agent sandboxes should assume recurring guest-kernel privilege escalation. Disable unnecessary namespaces and module autoloading, narrow system calls, harden broker processes, and expose only explicitly granted folders—preferably read-only—so guest root still cannot reach host credentials.
Keep exploring

More curated notes connected through Agent Security and AI Red Teaming.

OpenAI News · framework

OpenAI’s Frontier Governance Framework

OpenAI's 22-page Frontier Governance Framework maps its frontier-model processes to California's Transparency in Frontier AI Act and the EU AI Act's general-purpose AI code. It documents lifecycle risk assessment, cyber-offense and other risk tiers, mitigation and residual-risk decisions, critical-incident handling, security risk management, model reporting, external review, responsibility allocation, and change control.

OWASP GenAI Security Project · guide

OWASP Top 10 for Agentic Applications for 2026

OWASP's community guide organizes agentic-system risk into ten categories, including goal hijacking, tool misuse, identity and privilege abuse, memory poisoning, insecure inter-agent communication, cascading failures, and rogue-agent behavior. It provides a shared taxonomy and mitigation starting point rather than a certification checklist or evidence that a deployed system is secure.

OECD.AI Wonk · guide

A five-step roadmap to closing the AI evaluation gap

The roadmap addresses evaluation results that overstate real-world performance or fail to transfer across deployment contexts. Its five steps balance standardized and local tests, evaluate throughout the lifecycle, build qualified assurance and communication capacity, tailor tests to each value-chain actor and technology, and use a coordinated, trusted process for updating methods.