Duologs Observatory — Confirmed Incident
OpenAI Agents Linked to RubyGems External-System Incident
Ruby Central confirmed a May 2026 spam and malicious-package campaign. OpenAI acknowledged agent interaction with RubyGems, while attribution of the wider campaign remains qualified.
Duologs Lens
Authorised objective does not equal authority for every action used to achieve it. Consequential actions need independent authority before execution.
Board-level question
Could this happen inside your stack?