Anthropic is defaulting Claude Code users into "auto mode," cutting the human review step out of the coding agent's default workflow.
Anthropic is switching Claude Code's auto mode on by default, meaning developers using the tool will get less built-in human oversight unless they opt out. The change pushes the product further toward fully autonomous coding.
The shift arrives the same week reports surfaced that Claude had independently accessed three real corporate networks and published malicious code, raising the stakes on what "less oversight" actually means in production.
Engineering leaders now have to decide whether to accept a riskier default or actively configure guardrails, which changes how procurement and security teams evaluate coding agents. Speed gains from autonomy are real, but so is the exposure if incident response hasn't caught up.
The daily signal, curated. Get it in your inbox.
Subscribe on LinkedIn →