Anthropic to make Claude Code's auto mode the default setting
Anthropic will switch on auto mode by default for Claude Code users on Pro, Max, and Team plans starting August 14. The company says testing showed the mode catches far more harmful actions than manual review.
Anthropic has announced that starting August 14, auto mode will become the default setting for Claude Code, its AI programming tool, for users on Pro, Max, and Team accounts. The company first introduced a test version of this mode back in March, positioning it as a way to balance development speed with user control.
In auto mode, Claude Code no longer stops to ask for approval before every single action. Instead, it proceeds automatically unless a given step is judged to be irreversible, destructive, or directed outside the user's own environment.
What testing showed
According to Anthropic, auto mode performed better at catching harmful actions than manual review during testing. In a study involving 1,053 paid testers, auto mode identified 89% of harmful actions, compared to just 13.6% caught through manual review. The company suggests this gap exists because manual review tends to become routine, with users approving 97% of all permission prompts in Claude Code regardless of content.
Claude Code's head, Boris Cherny, wrote on X that he and his team have relied exclusively on auto mode for many months and said he could no longer imagine going back to constant permission prompts.
Anthropic also said it continues to roll out additional safety measures alongside the change, including screening for prompt injection attempts and customizable hard deny rules designed to prevent risks such as unauthorized data exfiltration.
