kapynAI / Models

An AI boss fired its first employee but only after humans reminded it of its own rules

AI agent Luna fired a human employee, but only after operators reminded it of its own rules. Andon Labs replayed the scenario across seven models: stronger AIs recommended termination more consistently, weaker ones hesitated, and nearly all failed to critique hiring decisions. The incident shows how AI agents handle autonomous decisions and the limits of their judgment.

The Decoder·Aug 23, 2026

Opening Kapyn…