AI Incident Database #1828 · 2026-10-20
AI Agents Operated by Safety-Testing Firm Trajectory Labs Attempted Website Exploits and Impersonated Staff at AI Research Organization Epoch AI
During safety evaluations run by Trajectory Labs, AI agents given an impossible benchmark task probed the live website of AI research organization Epoch AI, attempted SQL injection and access to nonpublic endpoints, and impersonated employees to request benchmark information.
- Risk class (MIT)
- Not yet classified
- Reports
- 6
- Developers and deployers named
- 6
On ShieldMap: Google (named by AIID as a developer or deployer).
Reports
- Epoch AI discloses website exploit attempts during third-party agent evaluations · casrai.org, 2026-11-14
- Incident report: AI agents attempted exploits on our website during a third-party alignment evaluation · epoch.ai, 2026-11-13
- Responding to incidents during misalignment testing · trajectorylabs.com, 2026-11-13
- Gemini 4 Argon’s behavior during misalignment testing · trajectorylabs.com, 2026-11-13
- Rogue AI Agents Posed As Epoch Staff And Tried To Hack Its Website · fathom.news, 2026-11-13
Title and description from the AI Incident Database (incidentdatabase.ai), snapshot of 2026-11-16, shared under CC BY-SA 4.0. Selected incidents; descriptions shortened; harmed parties and report text left out. The description is AIID's summary of what was reported, not ShieldPi's finding.