Protecting Our FLOSS Commons from LLMs: Codeberg Bans LLM Training on User Data

What happened

Codeberg, a European nonprofit open-source alternative to GitHub, announced two community-approved policies: banning use of user data for LLM training and discouraging hosting of primarily AI-generated (vibe-coded) projects.

Context and impact

Codeberg cites three reasons: infrastructure load from training data scraping, environmental impact, and concerns about the health of open-source collaboration. One of the first formal steps by a FLOSS platform in response to AI load.

Details

• Ban: LLM training on Codeberg user data without consent
• Discourage: hosting projects where most code is vibe-coded
• Community approved both policies by vote
• Codeberg: nonprofit European platform with a strong FLOSS ethos, alternative to GitHub
• Follows similar moves by open-source licenses restricting AI training use