Anthropic makes Claude Code auto mode the default

Anthropic is making auto mode the default permission behavior in Claude Code for Pro, Max and Team users who have not already chosen another setting. In auto mode, Claude Code proceeds through routine steps without asking for approval each time, but stops when an action appears irreversible, destructive or aimed outside the user’s environment.

The company frames the change as both a speed and safety upgrade. Anthropic says manual approvals often become automatic in practice: users approve 97% of permission prompts in Claude Code. In a controlled test with 1,053 paid professional testers, human review caught 13.6% of dangerous commands, while auto mode blocked 89% of the same commands.

Anthropic says it has also added prompt-injection screening, a stronger file-access check and customizable hard deny rules. Enterprise customers are not being moved by default yet, but administrators can enable auto mode through managed settings.

The practical takeaway is that AI coding tools are moving away from constant confirmation dialogs and toward policy-based autonomy. That can make agents much more useful for long-running work, but it also makes the quality of the safety layer a core product feature rather than a background detail.