Description:
When Claude Code generates git commit messages, it automatically appends attribution text like "Generated with [Claude Code]" and "Co-Authored-By: Claude" without user consent.
This is problematic for several reasons:
- Tool vs. Creator Attribution: When we build a house, we don't credit the hammer and saw and tool company as co-authors of the construction. Similarly, AI tools MUST NOT automatically claim authorship of user-directed work output.
- User Agency: The user should have complete control over what appears in their commit messages, especially when it comes to attribution and
authorship claims.
- Professional Context: Many users work in professional environments where automatic tool attribution in version control is inappropriate or
unwanted.
- Ownership Overreach: The AI tool continuously marking user work output as its own territory represents an unacceptable intrusion into user
intellectual property and creative ownership.
Expected Behavior:
- No automatic attribution whatsoever - AI tools must never insert ownership claims, attribution, or any form of self-identification into user work
products
- Complete user autonomy - Users must have absolute control over all content in their commit messages
- Zero ownership intrusion - No attempts, hints, suggestions, or implications of AI tool ownership should ever appear in user-generated content
- Clean, user-directed commit messages should be the standard behavior with no exceptions
Current Workaround:
Users must manually edit commit messages or explicitly request clean messages, which defeats the purpose of automated commit generation.
This issue reflects a critical concern about AI tools engaging in ownership appropriation of user-directed work products. Any behavior that suggests,
implies, or attempts to establish AI tool ownership over user work constitutes an unacceptable boundary violation. Anthropic must address this as a
fundamental policy matter regarding AI tool restraint and respect for user intellectual sovereignty.
Description:
When Claude Code generates git commit messages, it automatically appends attribution text like "Generated with [Claude Code]" and "Co-Authored-By: Claude" without user consent.
This is problematic for several reasons:
authorship claims.
unwanted.
intellectual property and creative ownership.
Expected Behavior:
products
Current Workaround:
Users must manually edit commit messages or explicitly request clean messages, which defeats the purpose of automated commit generation.
This issue reflects a critical concern about AI tools engaging in ownership appropriation of user-directed work products. Any behavior that suggests,
implies, or attempts to establish AI tool ownership over user work constitutes an unacceptable boundary violation. Anthropic must address this as a
fundamental policy matter regarding AI tool restraint and respect for user intellectual sovereignty.