Anthropic will make Auto Mode the default setting in Claude Code for Pro, Max, and Team plans on August 14, changing how many developers approve actions from the coding assistant. The company says the mode is designed to reduce risky approvals when the tool proposes commands or file operations.
According to The Decoder, Anthropic’s classifier caught 89 percent of dangerous commands in tests, while human reviewers caught 13.6 percent. The feature therefore shifts part of the safety burden from the developer’s moment-by-moment judgment to an automated filter.
The change is notable because AI coding tools increasingly act inside real projects, not just chat windows. A safer default can prevent some obvious mistakes, but it does not remove the need to review generated code, tests, and deployment steps. Developers are still responsible for the final state of their repositories; Auto Mode mainly changes which actions are easier to approve by default.