Zendesk never sleeps, and new features drop almost every week. We’re here to cut through the noise and give you the expert take on what actually matters. Here’s our analysis of the latest update from late October 2026.
Zendesk Finally Streamlines AI Permissions: A Sanity-Saving Update for Admins
What We Think:
This is a long-overdue and very welcome change for any organisation on an Enterprise plan using Zendesk’s AI features. Moving AI agent permissions into custom roles is a logical step that shifts management from a tedious, per-user task to a scalable, role-based strategy. Our verdict: A solid, necessary improvement that reduces administrative overhead and minimises the risk of human error. It’s a move towards a more mature and manageable platform.
The Reality Check: What’s Actually Changing?
Let’s cut through the official announcement lingo. For a long time, managing who could do what with AI agents was a separate, slightly clunky process. You had to go into an individual team member’s profile on the product roles page to set their access. If you had 50 agents, that was 50 individual settings to manage. It was inefficient and prone to error.
Now, for customers on Enterprise and Enterprise Plus plans, Zendesk is allowing you to manage Zendesk AI Access directly within your existing Custom Roles. This is an “opt-in” change, meaning it won’t happen automatically. An administrator has to go into each custom role and actively choose to manage AI permissions there. Once you opt-in for a specific role, that control is permanently moved to the main ‘Roles and permissions’ page for that role, away from the old individual profile settings.
For each custom role, you can now set one of four access levels:
- No access: The user can’t interact with AI agents. Simple.
- Client user: The most basic level, likely allowing interaction with existing AI agents.
- Client editor: This level probably allows for editing or refining AI agent configurations.
- Client admin: Full control over the creation, configuration, and management of AI agents.
If you don’t use custom roles, or you are on a lower plan, nothing changes for you. You’ll continue managing access the old way.
Our Expert Opinion: Why This Matters for Your Operations
The real win here is scalability and consistency. The old method was a classic example of unscalable administration. Imagine onboarding a new team of ten support agents. Previously, a Zendesk admin would have to set up their user accounts and then, as a separate step, go into each of the ten profiles to configure their AI permissions individually. It’s easy to miss one or set the wrong permission level.
With this change, the process becomes: assign the new agents to the “Tier 1 Support Agent” custom role, and you’re done. Their permissions for tickets, reporting, and now AI are all inherited from that single role definition. This centralises control and provides a single source of truth for who has access to what. It makes auditing permissions infinitely easier and reduces the risk of incorrect access being granted.
We also noted Zendesk’s comment that they plan to “add more granular AI agent permissions in a future release.” This is telling. It signals that what we have now is just the first step. While the current four levels are a huge improvement, we anticipate future updates will allow for more nuanced control, perhaps distinguishing between different types of AI agents or specific functions within the AI toolkit. For now, this new approach to Zendesk AI Access lays the essential groundwork.
This move is a clear sign that Zendesk is treating its AI suite as a core part of the platform, deserving of the same robust, role-based permission structure as other critical features. If your organisation is scaling its use of AI, this update is not just a nice-to-have; it’s a foundational requirement for good governance. For guidance on how to best structure your permissions, our team’s expertise in Zendesk implementation can help you build a model that scales with your business.
How We’d Handle the Rollout: An Action Plan
Here is what we’re advising our clients to do. Don’t just rush in and click “Opt in” everywhere. A planned migration will prevent any disruption. We recommend a phased approach to ensure a smooth transition to the new permission model for Zendesk AI Access.
| Step | Action | Why It’s Important |
|---|---|---|
| 1. Audit Current Access | Create a spreadsheet of all team members who currently have AI agent access. Note their name, current custom role, and their specific AI permission level. | This gives you a baseline. You need to know the current state before you can successfully migrate to the new one. |
| 2. Map to Roles | Analyse your audit. Do all agents in the “Customer Service” role need “Client editor” access? Or do you need to create a new custom role, like “AI Content Specialist”? Define the AI permission level for each of your existing custom roles. | This is the core strategic step. It ensures your role definitions accurately reflect your operational needs and security policies. |
| 3. Communicate the Change | Inform your Zendesk administrators about the new, centralised way of managing AI permissions. Explain that they should no longer use the individual profile pages for this. | Prevents confusion and stops admins from trying to use the old, now-defunct method for roles you’ve opted in. |
| 4. Pilot Test | Choose one non-critical custom role with a small number of users. Go to Admin Center > Roles and permissions, edit the role, and select “Opt in for AI agents”. Set the desired permission level. | A pilot test minimises risk. It allows you to verify the process works as expected before rolling it out to major teams like your entire support department. |
| 5. Verify and Roll Out | Ask a user in the pilot role to confirm their access is working correctly. Once confirmed, repeat the opt-in process for your other custom roles, one by one. | Verification is key. Once you’re confident, you can complete the migration for all relevant teams. |
| 6. Update Documentation | Update your internal runbooks and administrator guides to reflect the new, role-based procedure for managing AI access. | Good documentation ensures consistency and makes it easier to train new admins in the future. |
In conclusion, this update is a significant quality-of-life improvement for Zendesk administrators. By centralising Zendesk AI Access controls within custom roles, Zendesk is making the platform more secure, scalable, and easier to manage. If you’re on an Enterprise plan, we strongly recommend you adopt this new model following a careful and considered rollout plan. If you need a hand with the strategy or execution, don’t hesitate to contact your alboz. Solution Consultant.
