Anthropic sets Claude Code auto mode as default for Pro and Team plans

Anthropic has announced that starting August 14, Claude Code's "auto mode" will become the default setting for users on Pro, Max, and Team plans. This update is designed to enhance developer productivity and safety by reducing manual permission prompts during complex coding tasks.

To manage risks, the auto mode utilizes a classifier to monitor tool calls, specifically aiming to block harmful or irreversible actions. Anthropic reports that this system is more effective at identifying dangerous commands than human reviewers; in controlled studies, the auto mode blocked 89% of dangerous commands, while human testers caught only 13.6%.

Internal data suggests that teams utilizing auto mode ship an average of 25% more pull requests. Additionally, Anthropic noted that the extra token usage required by the classifier will not incur additional costs for users. While the feature is being rolled out to Pro and Team users, it remains opt-in for Enterprise users and those accessing Claude Code via API platforms such as AWS and Google Cloud, with broader deployment expected in the coming month.