Anthropic Makes Auto Mode the Default in Claude Code
In one line: Anthropic has reportedly made Auto Mode the default in Claude Code to reduce developer mistakes caused by bad approvals.
Key points
- Claude Code, Anthropic's coding assistant, has reportedly shifted its default behavior to Auto Mode.
- The stated goal is to protect developers from carelessly approving risky actions (bad approvals).
- By having the tool manage the approval flow by default, the change aims to cut both the fatigue of confirming every step and the risk of mistaken approvals.
Why it matters
As AI coding tools take on more real actions — editing files, running commands — the boundary between what is auto-allowed and what needs human confirmation has become a core safety control. Changing the default directly reshapes the workflow for a large base of developers.
Read more
How this story unfolded
- Anthropic's 'One Git Worktree Per Agent' Advice Clashes With Runtime Infra
- Anthropic Closes $965B Valuation Round, Files Stealth IPO — Claude Code Hits $2.5B ARR in Nine Months
- Anthropic Launches Claude Code Artifacts: Live Dashboards from AI Coding Sessions
- 'Agentjacking' Attack Uses Fake Bug Reports to Hijack AI Coding Agents