Anthropic will make Claude Code’s auto mode the default
Anthropic plans to enable auto mode by default in Claude Code for Pro, Max, and Team accounts starting August 14. The company introduced an experimental version in March as an option intended to combine faster execution with safeguards.
In this mode, Claude Code does not ask for approval before every action. It continues working unless an operation is judged irreversible, destructive, or directed beyond the user’s environment. Anthropic said testing with 1,053 paid participants found that auto mode identified 89% of harmful actions, compared with 13.6% identified through human review. The company noted that users approve 97% of Claude Code permission prompts, suggesting that repeated manual checks may become routine.
Boris Cherny, head of Claude Code, said he and his team had used auto mode exclusively for many months. Anthropic also reported adding safeguards including prompt-injection screening and customizable hard-deny rules designed to block actions such as data exfiltration.
Written by Sakhanda Wire AI from reporting by TechCrunch. Facts, quotes and full context are in the original article below.