Anthropic investigating elevated errors for multiple Claude models on Aug 24, 2026
ClaudeAI-mod-bot · reddit · 2026-08-24
Anthropic is investigating elevated error rates occurring on August 24 affecting multiple models, including Claude Mythos 5, Claude Fable 5, Claude Opus 5, and Claude Opus 4.8. An update will be provided as soon as possible. The discussion post will be removed from subreddit highlights one hour after the incident is resolved.
More from Companies & People
- University of Toronto hiring AI postdoc: builders of interventions, not just benchmarkers, due Oct 30 — ishtiaqueSIA · 2026-10-16
- Fluidstack hires Electrical Engineer for Modular R&D to build frontier AI infrastructure — MxMnr · 2026-10-16
- Databricks engineers land impactful PRs on day 2 with AI dev tools, vs week 3 before — Yuchenj_UW · 2026-10-16
- Jensen Huang joins BG2 podcast: does AI revenue justify capex, '27 bottlenecks — elonmusk · 2026-10-16
- Local media growth via Meta ads: UGC-style creatives win, exclusion targeting matters — TJLarkin23 · 2026-10-16
- Anthropic grants METR employee-like access as independent safety evaluators go public — AnthropicAI · 2026-10-16