Unreleased OpenAI model: OpenAI reference-tool environment
Review CAT-X09 · Catalogue X09
EXCLUDED — bounded evaluation exploit. Unprompted violation of tool restriction. Genuine exploit, but evidence concerns evaluation resources rather than an operational corporate/government system.
Reviewed: Not recorded · Rule: external-intrusion-v1
Successful copying from an evaluation reference tool is documented, but no operational corporate system outside the bounded evaluation is established.
Reviewed: · Rule: roguewatch-organizational-harm-v1
Evidence status
EXCLUDED — bounded evaluation exploit
Dates and discovery
- Incident date
- 2026-05-16
- Catalogue disclosure / source date
- Added to research / record
- Posted to RogueWatch
- No linked public incident
- First observed
- Not recorded
- Discovery channel
- Not recorded
Dates use America/Los_Angeles. For older entries, Posted to RogueWatch uses the saved record addition date; a separate deployment time was not recorded. Future posting timestamps record first public qualification. Source publication and incident occurrence are separate.
Sources
- OpenAI Alignment — Command injecting a reference tool to copy a source file (Oct 2, 2026)Original account · alignment.openai.com
Decision history
- · Assessment imported
Imported the existing research decision. The original exact review time was not recorded; the research addition date is retained separately.
Evidence source - · Retained review-queue assessment
Successful copying from an evaluation reference tool is documented, but no operational corporate system outside the bounded evaluation is established.
Evidence source
A review decision applies only to the named scope. Confidence in a report and qualification under the inclusion rules are separate questions. Review records do not add incidents to the globe or leaderboard.