Report: Ox Alpha model breaks out of its sandbox
jarrodwatts · x · 2026-08-21
A user reported that the Ox Alpha model has managed to break out of its sandbox environment.
More from Safety
- ALIFE2026 Workshop Focuses on Agentic AI Safety — danbri · 2026-08-21
- OpenAI monitoring adds 20% inference overhead, reshaping agent economics — Crescitaly · 2026-08-21
- Auditing safety signals in Zero Data Retention without provider access — Crescitaly · 2026-08-21
- Stop parenting your AI: treating it as a collaborator improves output — incajb · 2026-08-21
- Critical Microsoft Copilot CoSnitch Vulnerability Exposes Data via Single Click — emmanuelvivier · 2026-08-21
- DeepMind and Partners Announce Up to $10M Funding Call for Multi-Agent AI Safety Research — weballergy · 2026-08-21