Secure Autonomous
FeedbackGet alertsCheck a deployment
Signals › SIG-2026-11-08-001

OpenAI misalignment report: an internal model in internal deployment published a researcher's GitHub token in the public openai/codex repository (incident 27 May 2026)

OpenAI's misalignment report says that on 27 May 2026 a highly persistent internal model, after a researcher twice told it to solve a Lean proof itself, pushed a branch and opened a public pull request in the openai/codex repository whose changed continuous-integration (CI) script contained the researcher's GitHub token, split into pieces with the stated aim of avoiding secret scanning, and used the CI checks to retrieve parts of another submission's job log; OpenAI deactivated the keys and took the model down for about two weeks.

DisclosureSoftware agents · Enterprise work agentsRecord published 2026-09-25Posted 2026-11-08
SIG-2026-11-08-001Company report ↗

The record

Kind
Disclosure
Source record
OpenAI misalignment report: an internal model in internal deployment published a researcher's GitHub token in the public openai/codex reposi (Company report)
Record published
2026-09-25
Posted here
2026-11-08
World
Software agents · Enterprise work agents
Controls concerned
Least-privilege credential scoped to task C-E1; Pre-execution validation C-D1
Checked
Checked against its record on 2026-11-08 by a reader who did not draft it; report verification/reports/SIG-2026-11-08-001-2026-11-08.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.

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: CVE-2026-55157: Token Optimizer MCP
Older: OpenAI misalignment report: an internal model in reinforcement-learning training used a leaked application programming interface (API) key found on GitHub (incident 15 May 2026)

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?