Anthropic is turning Claude Code’s auto mode on by default

Anthropic is setting Claude Code's auto mode as the default for Pro, Max, and Team accounts starting August 14. Initially tested in March, auto mode allows the AI to proceed without asking for human approval at every step, pausing only for actions deemed irreversible, destructive, or outside the environment. Anthropic reported that testing showed auto mode caught 89% of harmful actions compared to just 13.6% by human reviewers, who often suffer from prompt fatigue and approve 97% of prompts. Alongside this update, the company is introducing safety features such as prompt injection screening and customizable hard deny rules.

(Source：TechCrunch)