Guardrails Service Disruption Impacting Customers

[identified] As of 20:07 UTC on December 2, 2025, our Engineering team has detected increased guardrail latency which may produce long response times for agents with attached guardrails. Engineering is working on a fix. We apologize for the inconvenience and will share an update once we have more information. Thank you for your patience and understanding as we work to resolve this issue.

[investigating] Our Engineering team is currently investigating increased guardrail latency, which may cause long response times for agents with attached guardrails. We apologize for the inconvenience and will share an update once we have more information. Thank you for your patience and understanding as we work to resolve this issue.

[monitoring] Our Engineering team has implemented a fix that has significantly reduced the impact of the issue, largely mitigating the outage. Although the situation has improved, some users may still experience intermittent latency issues. We are closely monitoring the results and will continue to work towards full resolution. We will provide another update once we have confirmed the issue is fully resolved.

[resolved] Our Engineering team has fully implemented and confirmed the effectiveness of the fix for the increased guardrail latency. System performance has returned to normal, and all affected services are operating as expected. We have observed stable metrics during extended monitoring and have not detected any further latency or interruptions. If you continue to experience problems, please open a ticket with our support team. We apologize for any inconvenience.