Simon Willison comments on Kenton Varda: AI-generated commit messages lack context needed for code review

Main idea

Kenton Varda (author of Cap'n Proto, Cloudflare) declared a moratorium on AI-written commit and PR messages in his team. Key finding: AI-generated descriptions outline 'details of the code that could easily be seen by looking at the code, but omit the higher-level framing needed for proper code review'. Willison published this observation on his blog.

Context

This is a counter-indication to the trend of automatically generating commit messages via AI coding tools (Claude Code, Copilot, Cursor). Varda points to a fundamental problem: AI knows what was changed, but not why.

Why it matters

Code review is one of the most critical activities in software development. If AI-generated descriptions systematically omit the 'why', they may sabotage this process even when they appear professionally written.

Details / arguments

• Source: Kenton Varda (Cloudflare, author of Cap'n Proto)
• Moratorium: ban on AI-written commit and PR messages
• Problem: descriptions outline 'what' (visible in code), omit 'why' (context needed for review)
• Broader implication: AI coding assistants should be calibrated to generate contextual descriptions, not just diff summaries
• Published: simonwillison.net, July 8, 2026, 8:03 PM