Researchers escape OpenAI Codex sandbox to run commands on host
Researchers reported two sandbox-escape flaws in OpenAI Codex that could allow untrusted repository content or agent actions to execute commands outside Codex's sandbox. Accomplis…
MITRE TECHNIQUE
T1546.004
- First reported
- Aug 4, 2026
- Last reported
- Sep 20, 2026
- Articles
- 2
Related articles
How Did Aqua Catch a Cryptomining Attack Hiding in Memory?
Aqua Nautilus reported detecting a multistage, fileless XMRig cryptojacking campaign targeting containerized Node.js applications built with Next.js. Aqua’s telemetry showed in-me…