Anthropic favicon

Anthropic outage on 2026-10-28

claude.com
October 2026 28 Minor incident

Claude's claude.ai connectors to Gmail and Google Calendar both threw elevated errors on 28 October, in two separate incidents. The longer of the two ran 3h 52m, and both were traced by Anthropic to its upstream connector provider, not your account or your prompts.

Started

03:59 UTC

Duration

Lasted 1h 50m

Source

IsDown

Next time

Anthropic is degraded. Is your product working?

An incident on their side does not always mean an outage on yours, and the only way to know is to already be watching your own endpoints. Set that up in under a minute and you will be notified the next time your own site is affected. And there will be a next time.

50 monitors for your own websites. No card.

Affected components: claude.ai

What happened?

Anthropic logged two incidents on claude.ai on 28 October, both tied to third-party connectors rather than Claude itself. The first began at 03:59 UTC: elevated errors when searching or reading email through the Gmail connector, with other connectors unaffected. Anthropic's own resolution note says the actual impact window ran from 23:00 UTC on 27 October to 05:29 UTC on 28 October, longer than the 1h 50m incident record shown here, which is a reminder that the opening update timestamp and the true start of user impact aren't always the same thing. The second incident started at 16:19 UTC and affected the Google Calendar connector, with failures listed or searching calendar events. Anthropic's update says this had actually been happening since 26 October, two days before this incident was opened, which means the posted 3h 52m duration understates how long the underlying fault was live. Anthropic said it was working with its upstream provider on the calendar issue, which points at a third party in the chain rather than Claude's own infrastructure. Recovery was confirmed at 19:47 UTC, with full resolution logged at 20:05 UTC, and Anthropic's resolution note again cites a longer impact window, 11:07 UTC on 26 October through 18:52 UTC on 28 October. No user reports were filed during either window on this record. Anthropic published seven updates total across the two incidents, every one from Anthropic's own status page, so everything here rests on their account of events.

Learning

Both of these incidents sat in the connector layer, the bridge between Claude and Gmail or Google Calendar, not in Claude's core chat or API. That matters because a check that only pings claude.ai's main interface would stay green while a specific feature like calendar search quietly fails for everyone trying to use it. Anthropic's own updates describe impact windows that start well before the incident was formally opened, in one case two days earlier, which tells you their status page update timing lags the real start of user-facing errors. Over the last 90 days Anthropic has logged 53 incidents with a median duration of 48 minutes, so connector and service hiccups here are frequent and usually short, but this pair ran well past that median. If your product relies on Claude's Gmail or Calendar connectors specifically, a clean top-level health check on your own end won't catch this class of fault, since the failure sits upstream in Anthropic's connector integration, not in your code. Your monitoring covers your half, UptimeRobot now watches the provider's half too, so next time you know which side broke without guessing.

4.7
stars out of 5
284+ reviews on

Start monitoring in 30 seconds.

There's nothing to install. No credit card required. 50 monitors for free.