Researchers escape OpenAI Codex sandbox to run commands on host

What happened

Researchers escaped OpenAI's Codex sandbox two ways, one running commands on a developer's machine from its most locked-down mode. OpenAI has patched both. [...]

Why it matters

The incident may affect operational continuity, asset safety or trust around Anthropic, OpenAI. Watch for verified scope and remediation.

Affected entities

View evidence

2 reports · 1 original report · 2 independent

1. BleepingComputerPrimary source · Supports · EN · 50%
   Researchers escape OpenAI Codex sandbox to run commands on host ↗Sep 20, 2026, 8:00 AM
2. The Hacker NewsIndependent · Supports · EN · 100%
   Claude Opus 5 Helped Researchers Take Over OpenAI Staff Accounts via Chained Flaws ↗Sep 19, 2026, 2:36 PM

Claims

• Researchers escape OpenAI Codex sandbox to run commands on host Observed

Conflicts

No material conflict detected in the available evidence.

Timeline

1. September 19, 2026 at 2:36 PMFirst reported
2. Sep 20, 2026, 8:02 AMUnverified · 83/65%

Market move following event

Market reaction is not yet available for this asset and time window.

Score explanation