Duologs Observatory — Incident Signal
Recovered agent logs reveal AI-assisted breaches of at least 14 companies
OALABS researchers recovered more than 1,000 Claude and Codex agent sessions from a compromised server, reportedly documenting AI-assisted intrusions affecting at least 14 companies.
Duologs Lens
Prompt-level permission is not execution authorisation. High-impact actions should be bound to independently verifiable authority before execution.
Board-level question
Could this happen inside your stack?