| Announced on | Rollout on |
| December 3, 2025 | December 3, 2025 |
We’re announcing an improvement to escalation configurations for advanced email AI agents.
This announcement includes the following topics:
What's changing?
Previously, advanced email AI agents could only handle one full use case before automation ended. This setting is now customizable, allowing you to define how many use cases an AI agent can complete before handing the conversation over to a human agent.
We’re also introducing the ability to add escalation blocks in the dialogue builder for advanced email AI agents. When a conversation reaches an escalation block, automation immediately ends and the conversation status is set to Escalated.
Why is Zendesk making this change?
You’ve shared that limiting automation to a single use case per conversation was too restrictive, especially when the use case was consumed by a simple knowledge reply and the AI agent could have continued assisting.
Additionally, conversations marked as Escalated previously required a human agent to reply within a short window of time. The new escalation block provides a more reliable way to flag these conversations for human attention.
What do I need to do?
No action is required. The default behavior remains unchanged until you decide to adjust it. For instructions, see Configuring the automation threshold for advanced email AI agents.
If you have feedback or questions related to this announcement, visit our community forum where we collect and manage customer product feedback. For general assistance with your Zendesk products, contact Zendesk customer support.