Summary
On March 12, 2026, between approximately 02:23 UTC and 05:50 UTC, some customers across all pods in the EU region experienced intermittent disruptions affecting multiple services. The cause was instability in the AI Agents chat-middleware, characterized by increased errors and frequent restarts, leading to slowed conversation handling, latency, and delays in chat and email automations.
Timeline
March 12, 2026 04:11 UTC | March 11, 2026 09:11 PM PT
We are investigating intermittent issues in Support (AI Agents and Agent Workspace) across multiple Pods (Europe region only). Chat conversations and email delays/errors are some of the symptoms observed at this stage. Our teams are investigating this issue at the highest priority. Next update in 30 minutes.
March 12, 2026 04:43 UTC | March 11, 2026 09:43 PM PT
We are continuing to investigate intermittent issues in Support (AI Agents and Agent Workspace) and Ultimate across multiple Pods in the Europe region. Customers are experiencing chat conversation and email delays or errors. We're also seeing dashboard loading errors. Engineering teams have applied resource and memory increases to the affected clusters, but service stability remains a work in progress. Thank you for your patience and understanding while we work through this. We will share further updates as soon as new information becomes available.
March 12, 2026 06:21 UTC | March 11, 2026 11:21 PM PT
We’re pleased to inform you that the intermittent issues affecting Support including AI Agents, Agent Workspace, and Ultimate in the Europe region have been mitigated. Our engineering team has deployed an optimization fix to the affected clusters aimed at improving stability. While this has had a positive impact, service reliability remains a work in progress as we continue to investigate the root cause. We recommend testing your environment to confirm everything is working as expected on your end. Please don’t hesitate to reach out if you experience any further issues or have any questions. Thank you for your patience and understanding throughout this process.
March 12, 2026 14:56 UTC | March 12, 2026, 07:56 AM PT
We’re pleased to confirm that the intermittent disruptions affecting Support services such as AI Agents, Agent Workspace, and Ultimate in the Europe region have been fully resolved. We appreciate your patience and cooperation as our team worked diligently to fix the issue.
Root Cause Analysis
The incident was caused by a configuration error in an AI Agent’s default reply referencing an inactive dialog, which triggered repeated processing between internal services. This led to resource overload and degraded system performance.
Resolution
To fix this issue, the team implemented a code fix that prevented the problematic bot behavior causing looping conversations. Following the deployment, error rates reduced, impacted conversations were resolved, and memory limits continue to be temporarily increased as a safety measure.
Remediation Items
- Prevent conversation loops from running indefinitely.
- Add a hard safety limit on automated actions per conversation.
- Improve isolation so one problematic bot or conversation can’t impact the whole region.
- Audit for similar loop risks across the platform.
- Strengthen detection and alerting for early warning signs.
- Fix and clarify on-call escalation paths.
FOR MORE INFORMATION
For current system status information about Zendesk and specific impacts to your account, visit our system status page. If you have additional questions about this incident, contact Zendesk customer support.