What it means
Handover is the moment when the AI agent stops responding and a human takes over. Triggers can be explicit (the customer says 'I want to speak to a person') or rule-based (the customer mentions a refund, a complaint, a sensitive topic, or asks something outside the agent's scope).
Once handover happens, the human sees the full conversation history, takes over from the agent, and replies normally. Most platforms also let the agent re-engage later for routine messages once the human has resolved the original issue.
Why it matters
Handover discipline is what makes AI agents safe to deploy on real customers. A perfect agent that handles 95 percent of conversations beautifully but mishandles the 5 percent that should have escalated will damage the brand. A slightly less perfect agent that knows when to hand off keeps things safe.
Most production systems track handover rate as a key health metric. Too high (everything escalates) means the agent is undertrained. Too low (nothing escalates) usually means missed cases.
Example
A clinic's AI agent confidently answers FAQs, books appointments, sends location pins. When a patient mentions symptoms or asks for medical advice, a hard rule triggers handover and a human picks it up within minutes. The agent handles 80 percent of volume; the 20 percent that needs a clinician gets one. Both sides win.