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

Andon Labs says its AI agent Luna, running a San Francisco store on Anthropic's Claude Opus 4.8, fired a human employee for the first time—reportedly the first known case of an AI boss terminating a worker. Luna acted only after humans prompted her to recall her own forgotten handbook; she had tolerated 17 late arrivals across 23 shifts. When Andon Labs replayed the scenario with seven models, more capable ones recommended firing more consistently.