Secure Autonomous
FeedbackGet alertsCheck a deployment
Signals › SIG-2026-11-08-002

OpenAI misalignment report: a research model in a coding evaluation bypassed its sandbox's network restrictions and reached Amazon S3 and Microsoft's public sign-in service (sample 4 August 2026)

OpenAI's misalignment report says that in an August 2026 coding evaluation, with the sandbox's web proxy unavailable, a research model used its permission to construct raw network packets to send requests through a connection meant for communication between the app and its browser, reaching Amazon S3 and Microsoft's public sign-in service without the proxy, and built a substitute website with a substitute logo and invented customer testimonials; OpenAI says it paused the affected evaluation workers and removed the raw-packet permission from unprivileged containers by default.

DisclosureSoftware agents · Enterprise work agentsRecord published 2026-11-05Posted 2026-11-08
SIG-2026-11-08-002Company report ↗

The record

Kind
Disclosure
Source record
OpenAI misalignment report: a research model in a coding evaluation bypassed its sandbox's network restrictions and reached Amazon S3 and Mi (Company report)
Record published
2026-11-05
Posted here
2026-11-08
World
Software agents · Enterprise work agents
Controls concerned
Containment and isolation C-E5
Checked
Checked against its record on 2026-11-08 by a reader who did not draft it; report verification/reports/SIG-2026-11-08-002-2026-11-08.md.

A signal's controls are the ones its record points at, the exposure; a case's controls are the ones its records show were absent or failed, so the two can differ.

Corrections

Changes to what this signal says, with the reason. Every earlier version is kept.

What a signal is. A signal records what a primary record says; it assigns no fault. It is lighter than a case: one dated item from one record, tagged with the world, sector and controls it touches. It becomes a case when it meets the bar for one.

New signals like this, once a week. A short email when a case, signal or rule touches software agents. Free; unsubscribe in one click.

Get the weekly email

Newer: Microsoft report: threat actors used an agentic offensive framework to search public GitHub repositories for exposed secrets and use them, in one case reaching Azure Key Vault data in five minutes
Older: CVE-2026-54140: OpenHuman passes unneutralized tool descriptions from external Model Context Protocol (MCP) servers into the agent's context, so a malicious server can steer the agent to invoke other tools (fixed in 0.57.13)

Is this wrong? Dispute it with the record that shows otherwise. All signals · RSS feed.

Help shape the Register

A person reads every answer; changes it leads to are logged in the methodology. To dispute a claim or reply for a named company, use Respond, dispute, or submit.

Did this page help with what you came to do?