If you run a Zendesk bot built with bot builder, answers or intents, this is the most consequential change to land in a Zendesk support stack this year.
Key Takeaways
For busy support leads:
- 1Three dates. Announced 23 June 2026. Development ends 31 August 2026. Removal 10 December 2026.
- 2Your bot stops working in December if you do nothing. This is an end of service, not a soft deprecation.
- 3You can rebuild now or wait for a tool. Zendesk says an automated, in-product migration tool is coming that will recreate an Essential agent without a manual rebuild, and tells customers they can wait for it. Details are due later in 2026.
- 4Your bill may move. Zendesk states that "because pricing is tied to outcomes, changes in the number of automated resolutions may impact your overall usage and associated costs."
- 5The upside is real. The replacement includes agentic capabilities that previously required the paid Advanced add-on.
The three dates that matter
Zendesk publishes these as a single table at the top of its announcement.
| Milestone | Date | What it means |
|---|---|---|
| Announced | 23 June 2026 | The deprecation was made public |
| Development ends | 31 August 2026 | Maintenance mode. Zendesk stops technical development except critical bug fixes and breaking-change support |
| Removal | 10 December 2026 | End of service. Any AI agents still on this technology stop functioning |
Zendesk recommends migrating before 31 August 2026 to avoid disruption. That recommendation is about avoiding a frozen product rather than anything breaking on that date. The hard cutoff is 10 December.
So you have roughly four months, not four weeks. Nobody should rush you with the August date, including us.
Who is affected and who is not
Affected: customers currently using AI agents Essential or legacy AI agent functionality. Zendesk names bot builder, autoreplies with articles and default messaging responses, and elsewhere includes answers and intents.
Not affected: customers who previously purchased the AI agents Advanced add-on.
The migration therefore falls on customers who did not buy the premium add-on. Zendesk has not published how that group breaks down by company size, so we will not guess.
Why Zendesk is doing this
Zendesk's stated reason is an upgrade. On 30 March 2026 it announced it was expanding access to its most advanced AI agent capabilities for all customers. This deprecation is the other half: retiring the Essential experience in favour of advanced functionality that Zendesk describes as built to support more complex automation and a more consistent customer experience.
That is a genuine improvement. Capabilities that sat behind a paid add-on, including agentic reasoning and multi-step procedures, become available more broadly.
The billing change worth reading twice
From the same announcement:
Because pricing is tied to outcomes, changes in the number of automated resolutions may impact your overall usage and associated costs.
The replacement AI is billed on automated resolutions. A more capable bot resolves more conversations, and more resolutions means a larger invoice. Zendesk deserves credit for putting that in writing rather than letting customers find it on a bill, but it does change how you should plan.
If you are budgeting, model your expected resolution volume after the upgrade rather than assuming your current bill carries over. Zendesk notes that existing automated-resolution overage configuration continues to be respected, so check yours before you migrate.
How much work is the migration, really
This is where an earlier version of this article was wrong, and it matters, so here is the correction.
Zendesk's migration guide describes a manual path today: you recreate the equivalent agent in the new experience, following steps that vary by configuration type. But the same guide adds that "in the future, an automated, in-product migration tool will be offered that will help you recreate an essential AI agent without manual rebuild", and tells customers they can "either follow the steps in this guide to unlock the potential of the new platform earlier, or wait for the tool to be available", with more detail due later in 2026.
So the honest position is that the size of this job depends on a ship date you do not control. If the tool arrives early, this is a small task. If it slips, you are rebuilding by hand against a December deadline.
What that does to switching costs
The usual reason teams stay on a support platform is not loyalty. It is that migrating is painful and existing flows work well enough. That inertia protects the incumbent.
For teams who rebuild by hand, that inertia largely goes away, because the work of reconstructing flows has to happen regardless. For teams who wait for the automated tool and get it in time, it does not.
Which means the useful question is not "should I switch" but "which of those two situations am I in". If you are going to rebuild manually anyway, evaluating alternatives costs you very little extra. If you can wait for the tool, staying is cheaper than it looks.
Your three options
Migrate inside Zendesk. The right call if you have deep Zendesk workflows, integrations, years of ticket history, or enterprise routing and audit requirements. You get better AI capability. Budget for a bill that moves with resolution volume.
Move to another traditional helpdesk. Help Scout, Front and Freshdesk are the usual comparisons. Expect to trade one usage-based AI meter for another, since outcome-linked AI billing is now the category norm rather than the exception, though the billing unit differs between vendors and most still charge per seat for assistant features on top. Check the current pricing page of any vendor you shortlist rather than relying on a comparison post, including this one.
Move to flat, AI-native pricing. This is what we sell, so weigh this paragraph accordingly. Corebee is a fixed monthly price with AI resolutions included rather than metered, so the number does not move when the bot gets better at its job.
The honest caveat is that we are not a like-for-like Zendesk replacement. If you rely on multi-tier routing, sandboxes, audit logs or a large integration catalogue, that depth is a real reason to stay. If your bot mainly deflects repetitive questions across chat, email and WhatsApp, it is not what you are paying for.
What we would do
Find out what you are actually on. If you bought the Advanced add-on, none of this applies to you.
Export before you touch anything. Get knowledge base articles, intent lists and conversation history out while the old system still runs. That keeps every option open regardless of what you decide.
Decide whether you are waiting for the tool. That single decision determines whether this is a small task or a project, and it is worth asking your Zendesk rep for a date rather than inferring one.
If flat pricing is the part that interests you, our Zendesk alternative page has the side-by-side, and you can try Corebee free without a card. For context on Zendesk's wider AI strategy, we also covered the Zendesk and Forethought acquisition.
Sources and corrections
Dates and quotations come from Zendesk's own announcement, "Announcing the end of support for AI agents Essential and legacy functionality: Important dates and migration guidance", its migration guide "Migrating to the new AI agents experience", and its 30 March 2026 announcement on expanding AI agent access, all on support.zendesk.com.
Two corrections to our earlier coverage. In May 2026 we published a piece describing an 11 May shutdown of Essential AI. 11 May was the start of a phased rollout that ran to 12 June 2026, during which advanced capabilities moved into Suite and Support plans. It was not the shutdown, and the end of service is 10 December 2026. Separately, an earlier version of this article said migration was manual with no automatic path. Zendesk's migration guide says an automated in-product tool is coming, and that sentence has been corrected above.