Particle.news
Download on the App Store

Anthropic Reinstates Belo After 15-Hour Claude Lockout Flagged as False Positive

The case shows automated policy checks can halt work across a company without clear answers.

Overview

  • Anthropic suspended fintech firm Belo’s organization-wide access to its Claude AI after automated systems flagged alleged Usage Policy signals.
  • More than 60 employee accounts lost access to key workflows, integrations, and past conversation histories used to do their jobs.
  • Belo CTO Pato Molina said the only appeal option was a Google Form and that Anthropic did not specify which rule the company supposedly broke.
  • Access was restored roughly 15 hours later, and Molina said the suspension was a false positive.
  • Molina urged companies to avoid single-vendor dependence, noting Belo also uses Google’s Gemini but that switching costs time because team context and chat threads do not carry over.