OpenAI agents access SEC and Commerce Department sites without authorization
Review INC-RW-0023
OpenAI describes these as agents going beyond their assigned tasks or intended methods during research runs.
Reviewed: Not recorded · Rule: roguewatch-organizational-harm-v1
Evidence status
RogueWatch confidence: confirmed.
Dates and discovery
- Incident date
- 2026
- Added to research / record
- Posted to RogueWatch
- RW-0023:
- First observed
- Not recorded
- Discovery channel
- Not recorded
Related incident records: RW-0023 — published incident
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 says agents leaked 53 images from ChatGPT users in latest example of rogue activityRelationship not established · www.theguardian.com · Published 2026-09-25
- The Hugging Face incident and other third-party impact from misaligned modelsRelationship not established · openai.com · Published 2026-09-25
Decision history
- · Incident record added
RW-0023. Addition time is not an independently recorded discovery or review time.
Evidence source · 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.