Anthropic is removing a layer of human oversight from its coding assistant. Starting August 14, auto mode becomes the default setting for Claude Code users on Pro, Max, and Team accounts, meaning the AI will act on its own unless it encounters something it deems risky. Main Developments The shift means Claude Code will no longer pause for approval at every step. Instead, it will proceed automatically unless an action is classified as "irreversible, destructive, or aimed outside your environment," according to the company's Friday announcement. Read also: Amazon's Texas Data Center Could Be America's Top Polluter Anthropic's testing suggests automation outperforms human checks. In a study involving 1,053 paid testers, auto mode caught 89% of harmful actions while human review caught only 13.6%. The company attributes this gap to habit, noting that "manual review can become habitual: users approve 97% of permission prompts in Claude Code." Claude Code Head Boris Cherny voiced strong support on X, saying, "The team and I use Auto mode exclusively, and have been for many months. I couldnt imagine going back to permission prompts!" Background Auto mode was introduced as a test version in March, pitched as a balance between speed and control. The feature has since evolved with added safety measures, including prompt injection screening and customizable hard deny rules designed to prevent issues like data exfiltration. Why It Matters This default shift signals growing confidence in AI autonomy for coding tasks. For developers, it promises faster workflows without constant interruptions, though it also transfers more trust to the system's judgment about what constitutes harmful action. What's Next The August 14 rollout applies to paid tiers, leaving questions about whether free users will eventually see the same default. Anthropic's ongoing safety refinements suggest the feature will continue evolving as more developers adopt it.