News
Claude Code Auto Mode Becomes the Default on August 14 — Anthropic's Own Study Says Humans Catch 13.6% of Harmful Actions and the Classifier Catches 89%
Anthropic confirmed auto mode ships as the default for Pro, Max and Team plans starting August 14, replacing per-call approval prompts with a classifier that inspects each tool call for irreversible, destructive or out-of-bounds actions. The justification is a study of 1,053 paid testers where auto mode caught 89% of harmful actions versus 13.6% for human review — alongside the finding that users already rubber-stamp 97% of permission prompts. Anthropic also stopped charging Pro/Max/Team users for classifier overhead effective immediately; Enterprise, the API, Bedrock, Google Cloud Agent Platform and Microsoft Foundry stay opt-in for now.
Source
↳ Follow the thread