News
Anthropic says Claude writes and reviews 90% of its team’s code. The catch is how it checks the work.
The New Stack · By Amanda Caswell ·
The Brief
- Anthropic says Claude writes and reviews roughly 90% of the code produced by the team building Claude Code.
- Anthropic describes how it verifies this work without having engineers inspect every change.
- The details come from a Monday engineering post by an Anthropic technical staff member, as reported by The New Stack.
Why it matters
It shows one concrete approach to verifying largely AI-written code without manual review of every change, which is useful for teams weighing similar workflows.
Who’s affected
- Teams using AI agents to write and review code
- #ai-coding
- #code-review
- #claude-code
- #verification
- #anthropic
Anthropic says Claude writes and reviews roughly 90% of the code produced by the team building Claude Code, and the company has spelled out how it verifies that work without making engineers inspect every change. In a Monday engineering post , Anthropic technical staff member…