Cases › operational mistake › LR-2026-0638
Claude global outage causes elevated errors and latency across all models
Preliminary Why: 1 report: Cyber Press · subject to change · 1 independent report
Status: action reported, no fix
No source states: loss or records, agent tool, could do, deployer, industry.
Severity moderate: Test: moderate, systems of others touched with a bounded scope stated. Fact: the disruption affected third-party developers, enterprises and users, and the incident was formally closed at 22:36 UTC.
Anthropic’s Claude AI platform, covering claude.ai, the Claude API, Claude Code and Claude Cowork, suffered a service disruption on 29 July 2026 with elevated error rates and latency across all models. Anthropic’s engineering team identified the source of the errors by 20:33 UTC and confirmed the problem spanned multiple models rather than a single deployment. At 21:38 UTC recovery efforts were underway and most models were improving. At 22:20 UTC Anthropic reported recovery across all models and normal success rates. The incident was formally closed at 22:36 UTC after roughly two hours and 47 minutes. Anthropic has not published a detailed post-incident report or root cause analysis, so the cause is not stated in the sources.
What happened
What was done
Risks
Primary63 incidents
Rising: 5 vs 2.2 a week
Sorted by rule from the case's fields; not yet read from the sources.
Controls
Not yet read for controls. Controls that mitigate the primary risk:
HarnessTool allow-list · Fail-closed defaults · Stop control
ApplicationHuman approval for high-impact actions · Change review and separation of duties · Least-privilege access for the agent · Production kept apart from test · Activity logs · Adversarial testing before deployment
Infrastructure / targetSimilar incidents
Evidence
No source states: what the AI was doing, what it reached or did, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.
Show the full account: 5 items from the primary sources
What happened, in order
- 2026-07-29 19:49 UTC Outage began; status page flagged the issue as under investigation.
- 2026-07-29 20:33 UTC Engineering team identified the source of the elevated errors across multiple models.
- 2026-07-29 21:38 UTC Update issued: recovery efforts underway and most models beginning to improve.
- 2026-07-29 22:20 UTC Incident moved to monitoring status; recovery reported across all models.
- 2026-07-29 22:36 UTC Incident formally closed after monitoring for recurrence.
Show every field, with its source and grade
The AI system
What happened
Impact
Notes on gaps
The press article's infrastructure and load-balancing explanations are speculation and were not recorded as facts. Counts of affected parties, users, records and losses are not stated. The 19:45 to 21:26 UTC window given by the press is a different window from the stated 19:49 onset and was not used. Dates and times are stated in the source; no dates were estimated.
LR-2026-0638 · version 1 · published 2026-11-01 after automatic checks; not yet checked field by field by a person · JSON