Duo Security Outage History

The Duo Security outage affected US deployments during sign-ins using Microsoft Entra MFA, with users seeing 504 gateway timeouts when authenticating. The issue began around 2:06 PM UTC and quickly manifested as widespread login failures across the United States, with a notable surge in user reports around 3:00 PM UTC indicating broad impact across multiple regions. The disruption impacted core authentication functionality and related Phone Call Delivery components, hindering access to apps that rely on Duo for multi-factor authentication.

Duo's status updates identified Entra MFA gateway timeouts and associated delivery issues, with a fix deployed around 4:40 PM UTC and ongoing monitoring thereafter. The outage lasted roughly 3:05 PM UTC to 4:40 PM UTC (about 1 hour 35 minutes) before services began to stabilize. By around 6:20 PM UTC many core authentication and Phone Call Delivery components were reported as operational again, with continued checks to ensure full restoration across all affected deployments.

Users attempting to authenticate with Duo Security during the DUO4 deployment faced authentication failures, impacting login flows that rely on Duo for multi-factor authentication and affecting access to services protected by Duo 2FA. The incident began at 6:40 PM UTC and was resolved by 8:20 PM UTC, lasting about 1 hour and 40 minutes. The provider issued 2 official status updates during the incident as teams investigated and then restored normal operation.

Duo Security experienced a disruption in its Support Phone System, causing issues with call connectivity between customers and Duo Support Engineers. This affected the Duo Support / Phone Support channel and temporarily hindered phone-based contacts. Duo advised customers to use alternate channels, such as email (support@duo.com) and the Support Portal, while the issue was investigated. The outage began at 7:40 AM UTC and was resolved by 9:00 AM UTC.

From Oct 20, 2025 7:55 AM UTC, Duo Security experienced email delivery issues across all deployments, affecting email-based communications and related authentication workflows. The disruption expanded by Oct 20, 2025 8:35 AM UTC to include phone call and SMS delivery for authentication, with a large set of Duo Admin Panel components showing degraded behavior. The incident prompted five status updates and led to a broad degradation affecting dozens of Admin Panel components. The outage persisted for most of Oct 20 and into Oct 21, with recovery beginning around Oct 21, 2025 3:00 AM UTC. User reports from Denver and Knoxville described login push delays and related access issues during the event.

Duo Security experienced an authentication outage in which Duo Desktop was not recognized during authentications on Windows 11 25H2, causing sign-in failures for users relying on Duo for MFA. The issue affected multiple core authentication services and the Admin Panel, impacting a broad set of Duo components and preventing successful logins. The outage began at 12:20 PM UTC and was resolved by 8:06 PM UTC, lasting approximately 7 hours and 46 minutes. The incident included two official status updates documenting the investigation and restoration steps, and listed the affected components on the status page. Duo recommended avoiding the Windows 11 25H2 upgrade and, if already upgraded, downgrading to the previous version to resolve the problem.

Duo Security reported a full deployment outage on its DUO34 deployment, taking all services offline. The outage impacted 5 components: DUO34 / Phone Call Delivery, DUO34 / SMS Message Delivery, DUO34 / Admin Panel, DUO34 / Core Authentication Service, and DUO34 / Push Delivery. Users were unable to authenticate or access related services during this window. The outage began at 7:15 PM UTC and was resolved by 8:00 PM UTC, lasting roughly 45 minutes.

The provider issued 2 status updates over the course of the incident, indicating a full outage on the DUO34 deployment and later confirming restoration of all components to normal. By the end of the window, all listed components were back to up.

An Entra ID sync issue affected Duo Security's European deployments, causing custom attributes to fail syncing from Entra ID to Duo. This degraded four Admin Panel components (DUO3, DUO38, DUO47, DUO71) while standard attributes continued to sync. The incident began with an investigation at Sep 29, 2025 9:50 AM UTC and was limited to European deployments.

The provider posted two status updates during the incident, and the issue was resolved by Sep 29, 2025 5:35 PM UTC. The incident lasted roughly 7 hours and 45 minutes, with affected components returning to normal once the issue was resolved. Geographic scope remained Europe.

The Duo Security AI Assistant in the Admin Panel became inaccessible for administrators on new Duo accounts, impacting onboarding and admin workflows that rely on AI guidance. The issue affected AI Assistant components across multiple admin configurations, while accounts created before today were unaffected. The outage began on 2:56 PM UTC and was resolved by 9:45 PM UTC, lasting about 6 hours and 49 minutes.

Duo Security posted two status updates during the incident as they investigated and deployed a fix, and access to the AI Assistant returned to normal by 9:45 PM UTC across affected admin configurations.

Duo Security reported an outage affecting Admin Panel login for cross-deployment logins across dozens of Admin Panel deployments. The incident began at Sep 9, 2025 4:45 PM UTC and was resolved by Sep 9, 2025 5:10 PM UTC, lasting about 25 minutes. Admins could experience login failures during this period. Duo Security's status page noted the issue as Identified and that teams were working to correct it; the incident included two official updates, and by Sep 9, 2025 5:10 PM UTC the affected deployments had returned to normal.

Duo Security experienced a short degraded state affecting updates or changes to users in the Admin Panel and related SSO functions across 12 components on multiple deployments. Users attempting to manage accounts or perform Admin Panel operations may have faced delays or incomplete changes during the incident. The outage began on Aug 26, 2025 2:46 PM UTC and was resolved by Aug 26, 2025 3:00 PM UTC, lasting about 14 minutes. The status page indicated an initial investigation and later restoration, with 2 official updates issued during the incident.