Cursor outage: Grok Bot computers unreachable for some users

Affected components

• Grok Bot

1. resolvedSep 7, 2026, 21:37 UTC

   This incident has been resolved.

2. monitoringSep 7, 2026, 20:41 UTC

   A fix has been implemented and we are monitoring for service restoration.

3. identifiedSep 7, 2026, 20:33 UTC

   The root cause has been identified and a fix is being implemented.

4. investigatingSep 7, 2026, 20:32 UTC

   We have identified an issue affecting Grok Bot. Some users may be unable to reach their Grok Bot computer, and Recover or Reset may fail. Your data is not affected. A fix is rolling out and we will post another update shortly.

What APIStatus.watch saw

Our 5-minute checks didn't record a change in Cursor's overall status around this incident. Smaller or regional incidents often leave a provider's overall status green.

Other services with incidents at the same time

Overlapping incidents aren't necessarily related.

• Elasticsearch: Replication bug causing slow recoveriesMajor · Sep 1, 2026, 08:16 UTC
• MongoDB Atlas: Impaired Cluster operations in AWS me-central-1 and AWS me-south-1Major · Sep 3, 2026, 16:44 UTC
• OVHcloud: [EU-SOUTH-MIL][Containers & Orchestration] - Managed Private Registry incident notificationMinor · Aug 26, 2026, 15:00 UTC
• Snowflake: INC20000150Minor · Aug 28, 2026, 00:07 UTC
• Telnyx: Intermittent delays for some outbound messagesMinor · Sep 2, 2026, 21:33 UTC

How often does Cursor have incidents?

Cursor reported 50 incidents in the last 90 days, 15 of them major or critical. A typical incident lasted 1h 8m. See Cursor's uptime and incident history