Stop automatic ai replies after takeover

Once a human operator has explicitly taken over an AI-held conversation, the AI never auto-sends another reply to it.

When this happens: A new customer message arrives after a human take-over.

Step by step

1
Takes over an AI-held conversation. ResponsiblePartyType becomes User.
EasyCRM sends another message. The Auto Reply gate checks the responsible party; it is User, not Ai/Unassigned, so the gate fails and no AI reply is sent — the operator handles it.

Other paths

Exception — conversation resolves and reopens or restarts

The conversation later reaches a terminal status and, per the reopen/new-conversation rule (FR-14, UC-14.7/UC-14.8), a later customer message either reopens it or starts a new one. The Auto Reply gate is re-evaluated from the start on that reopened or new conversation, exactly as for any inbound message — the earlier human take-over does not permanently exclude the customer from Auto Reply on a later, distinct conversation episode. This is the configured exception referenced by BRD §3.9 (unless the configured rules provide otherwise).

Operator returns the conversation to the ai

Once the complaint is settled, the operator clicks Return to AI (UC-24.13 AF, FR-24-014). Sets the AI as responsible party; from the next inbound message the Auto Reply gate applies again as for any AI-held conversation.

Good to know

Automatic AI replies stop once an operator has taken over, unless the reopen/new-conversation rules restart the episode (BRD §3.9).
Copilot remains available to the operator on a human-held conversation regardless — this rule only governs *automatic* sending.