video1.0<iframe src="https://www.loom.com/embed/ab123c6285154ab39498db1d03e4fa9b" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/ab123c6285154ab39498db1d03e4fa9b-db8ec760a51d0e07.gif280.512This Loom explains the human escalation beta feature for AI responses and how to configure it. It describes using rules based on a sensitivity setting (low, moderate, high) to decide when the AI should stop and allow a team to take over. Viewers can customize the escalation prompt and choose whether the AI sends a final acknowledgement message before escalating, plus set a fallback message for unexpected errors. It also covers how to test scenarios and what happens when escalation triggers, including where notifications appear in the inbox and via email and the team options when an escalation is open or resolved.