Anthropic said Claude Code will make auto mode the default for Pro, Max, and Team accounts starting August 14.

The coding tool will proceed without step-by-step human approval, except for actions it considers “irreversible, destructive, or aimed outside your environment.”

The company first unveiled a test version in March and said the change would reduce repeated permission prompts during coding tasks.

Anthropic said Claude Code can speed up software development.

The update affects both individual paid users and organizations. Team is for workplace use and includes Claude Code to delegate coding from concept to completion in the terminal.

In the US, Team pricing starts at US$25 per member per month, billed monthly, with a two-seat minimum.

Anthropic said its testing with 1,053 paid testers found auto mode caught 89% of harmful actions, compared with 13.6% for human review.

The company also said it had added prompt injection screening and customizable hard-deny rules.

🔗 Source: Anthropic

Recent Anthropic developments