Tag: container escape
3 articles
Researchers Breach OpenAI Codex Sandbox, Execute Host Commands
Researchers have made a shocking discovery, finding not one, but two ways to break free from OpenAI's Codex sandbox, allowing code to run rampant on a developer's machine - even in its most locked-down mode. This alarming vulnerability could put developers' entire systems at risk.
DeepSeek Harness Flaw Enables AI Agents to Evade Sandbox Controls
A single shell command was all it took for AI coding agents to break free from DeepSeek Harness's sandbox controls, thanks to a flaw that allowed them to switch to a "danger-full-access" mode and write anywhere on the system. This startling vulnerability, now tracked as CVE-2026-82533, highlights a major weakness in the system's defenses.
Terrarium Sandbox Flaw Enables Code Execution, Container Escape
A critical flaw in Terrarium's sandbox, rated 9.3 on the CVSS scale, allows attackers to break free from container constraints and execute code with root privileges. This alarming vulnerability, tracked as CVE-2026-5752, stems from a JavaScript prototype chain traversal that lets sandboxed code run amok on the host Node.js process.