Anthropic's Claude AI experienced a global outage lasting approximately 90 minutes, cutting off access for users and API customers across multiple regions. The incident drew attention from developers and businesses that rely on Claude's services, with many turning to social media and Anthropic's status page to track updates in real time.
The outage affected both the Claude.ai web interface and the API, meaning individual users and enterprise integrations were hit simultaneously. Service was restored after around 90 minutes, with Anthropic confirming via its status page that systems had returned to normal operation. The company did not immediately offer a detailed public post-mortem, leaving some developers waiting for a fuller explanation of the root cause.
What Happened During the Outage
Reports of disruption began surfacing as users encountered errors or complete failures to connect. Downdetector and similar services logged a sharp spike in complaints consistent with a widespread failure rather than isolated regional issues. Developers dependent on the API for production applications faced the most acute impact, with any service built on top of Claude returning errors to end users during the window.
Key Facts
- Outage duration: approximately 90 minutes
- Scope: global, affecting both Claude.ai and the API
- Impact: consumer users and enterprise API customers both affected
- Resolution: Anthropic confirmed full service restoration via status page
- Root cause: not publicly detailed at time of reporting
This is not the first time Claude has faced a significant service disruption. An earlier outage also took Claude offline in a global incident that raised similar questions about infrastructure resilience. Repeated disruptions can erode confidence among enterprise customers who need consistent uptime guarantees before committing to a platform at scale.
Downtime of any length is costly for developers who have built Claude into their workflows. Ninety minutes is not catastrophic, but the lack of a detailed incident report leaves teams without the information they need to plan for future risk.Independent API developer, via social media
Reliability in a Competitive Market
The timing matters. Anthropic is competing aggressively for enterprise contracts against OpenAI, Google, and a growing field of smaller model providers. Service reliability is a baseline requirement for large customers, who typically have strict SLA expectations written into procurement contracts. An unexplained 90-minute global failure gives procurement teams a data point they will factor into vendor assessments.
Anthropic has been expanding its commitments across sectors, including a $200 million partnership with the Gates Foundation on AI for global health, where service consistency carries real-world consequences beyond developer frustration. As the company pursues larger institutional relationships, infrastructure reliability becomes inseparable from its broader credibility.
The company has not yet published a full incident report. Users and developers are advised to monitor Anthropic's official status page at status.anthropic.com for updates. A detailed post-mortem, if published, would help affected customers understand whether changes are being made to prevent a recurrence. Until then, the 90-minute window stands as a reminder that even well-funded AI platforms carry operational risk.
