Anthropic announced that auto mode, which uses a classifier to automatically block potentially dangerous commands rather than prompting the user for approval, is becoming the default operating mode for Claude Code on Pro, Max, and Team plans starting August 14, 2026. The company reports that in its testing the auto-mode classifier caught 89% of dangerous commands, compared to just 13.6% caught by human reviewers manually approving actions. Anthropic plans to extend the default to Enterprise and API users over the following month and is removing the token charges previously associated with running the classifier for paid plans.
