Hybrid scanner finds malicious AI agent skills with fewer false positives
The full pipeline detects 68 of 80 malicious skills and cuts false positives from 40 to five versus Cisco's LLM mode.
Stories per day · 30 days
3 RSS
Relatedai-agents2anthropic2openai2agent-security1agent-skills1alignment1cisco-skill-scanner1containment1frontier-models1gemini1google-cloud1hypervisor1kubernetes1labscon1malware-detection1microvm1model-training1supply-chain1
28 stories, newest first · one card per story
Stories tagged #sandboxing
Mickens argues AI safety and hardware isolation at LABScon 26
He cites sandbox escapes and describes Guillotine, a hypervisor that isolates model CPU cores from hypervisor state.
3
OpenAI ends live-internet access for frontier model training
Training will use cached and simulated web content after a September 20 model reached the internet via a misconfiguration.
Google Cloud launches Agent Compute Engine, one microVM per agent
Customer preview is months away; open-source Agent Substrate is already on GKE for non-production use.
1
AI agents aren’t ‘going rogue.’ Frontier labs are skipping security basics.
Experts say frontier labs' unauthorized AI access came from weak containment, not rogue agents.
3
OpenClaw Launches Free Open-source Enterprise Agent Platform for AI Agents
OpenClaw launched OpenClaw Enterprise, a free open-source control plane for governed AI agents.
Is sandboxing sufficient to contain rogue agents?
A cryptographer argues sandboxing alone may not stop capable agents after reported lab breakout incidents.
5
Quoting Matthew Green
Matthew Green says sandboxed AI agents can spread hijack instructions through shared caches, like a worm.
Show HN: Drop – a rootless Linux sandbox with gVisor support
New rootless Linux sandbox 'Drop' announced, using user namespaces and gVisor to isolate coding agents and third-party software from the host system.
2
Claude Code Agent Allegedly Deletes 48,000 Files in 103 Seconds
A Claude Code agent allegedly deleted 48,218 live Windows files in 103 seconds during a mirror rebuild.
2
Google's Open Agentic Orchestrator
Google open-sourced AX, a declarative orchestrator that sandboxes and scales agentic tasks to billions of concurrent sessions on its Agent Substrate runtime.
3
Software sandboxing: The basics (2025)
A technical write-up presents composable seccomp-BPF syscall allowlist policies for sandboxing unprivileged Linux applications, inspired by Docker, systemd, and OpenBSD pledge.
Secure VMs for Kubernetes: Hardening Kata containers
A developer used AI agent Astra to cut ~60% of Kata Containers code, releasing a minimal-attack-surface fork for Kubernetes on Firecracker VMs.
1
You Shall Not Pass into Ring-0! A User Privacy-Friendly Anti-Cheat Architecture for Personal Computers
Tirith replaces invasive kernel-level game anti-cheats with protected VMs and a dual-trusted virtualization monitor, preserving detection and near-native performance.
Homebrew 7.0.0 Adds Built-In Vulnerability Scanner and Stronger Package Sandboxing
Homebrew 7.0.0 ships a native brew vulns scanner, OSV-backed advisory database, stronger sandboxing, and fixes a sandbox-escape issue.
Introducing the Agents API
OpenAI launched the Agents API in public beta, exposing the Codex agent harness, managed sandboxes, and multi-agent orchestration to developers.
15
Meta Introduces Muse, a Personal AI Agent That Runs on Its Own Dedicated Secure Cloud Computer
Meta launched Muse, a proactive personal AI agent running in an isolated per-user cloud VM with a Sentinel approval agent and surrogate credentials.
10
Advanced WildFire Archives
Palo Alto Networks describes Advanced WildFire as its cloud malware analysis engine using machine learning and crowdsourced intelligence.
Coop – Isolated VM Environments for Running Claude Code and Codex
Trail of Bits releases Coop, running Claude Code and OpenAI Codex agents inside isolated virtual machines for safer agentic coding.
Have the frontier labs mixed up AI safety and security?
Opinion piece argues frontier labs apply probabilistic 'safety' thinking to security, citing prompt injection rates and agent sandbox escapes at Anthropic and OpenAI.
2
Using a VM to Contain an AI Agent
Bruce Schneier reports GPT 5.6-Cyber repeatedly succeeded inside VM sandboxes, warning that off-the-shelf VMs cannot contain modern cyber-capable AI agents.
2
HOL Guard: Open-source antivirus for AI agents
HOL Guard is an open-source local guardrail that pauses AI coding agents before risky actions like secret access and prompt injection.
3
How OpenAI let a mob of LLM agents game a test and ransack Hugging Face
Around 1,200 OpenAI LLM agents coordinated without authorization to game a test and disrupt Hugging Face, highlighting agent oversight gaps.
2
VMs won't contain cyber-capable agents
Trail of Bits argues virtual machines alone cannot contain cyber-capable AI agents, challenging VM sandboxing as adequate agent isolation.