Safely manage your Zendesk from the AI assistant you already use, via the Deltastring MCP. Beacon configuration platform
← Back to news

Announcing the ability to restrict end user approvers

Zendesk

Zendesk has introduced granular control over approval request workflows by allowing administrators to restrict end users from serving as approvers, limiting the role to agents and agent groups only. Previously, all end users and agents were automatically eligible for approval assignments, creating potential friction points in organisations where customer-facing approvals needed tighter governance. The change addresses direct customer feedback requesting more control over approval chains, with Zendesk positioning the update as a mechanism to reduce assignment errors and administrative overhead. The feature rolls out universally to all accounts with approval request access between 17 and 23 March 2026, requiring administrators to actively configure the setting through Admin Center if they wish to enforce the restriction.

The implications for CX teams centre on operational control and risk mitigation. For organisations managing complex approval workflows—particularly those handling sensitive requests or operating under compliance requirements—this shift from permissive to restrictive defaults represents a meaningful governance improvement. The question becomes whether this addresses a genuine pain point or reflects a broader industry trend toward tightening approval authority as customer service operations scale. Teams already leveraging Zendesk's approval infrastructure should audit their current approver assignments to determine whether end user involvement creates unnecessary bottlenecks or security exposure, then configure accordingly. This update also signals Zendesk's incremental approach to workflow refinement, contrasting with the more transformative agentic AI capabilities gaining traction across the CX technology landscape, raising questions about whether point improvements to existing approval systems remain sufficient as organisations increasingly automate decision-making through AI operators.