Workshop: On the Internet, Nobody Knows You’re a Bot
summary
Group settles on concrete problems: contributions generated with Gemini or DeepSeek-like tools must still be submitted by a human and remain under CLA covers AI contributions, so copyright and patent grants still apply; projects debate allowing bots to auto-fix typos or LGTM-only changes versus granting approve rights. Proposals include staged trust (run an AI for six months on dependency bumps or tests and then flip a switch — bot auto-approve after 6 months), scan-for-plagiarism tooling, and drafting a CNCF-wide policy.
AI-written summary · about
chapters (8)
- 0:00AI in open source projects
- 1:45Legal constraints with CLA
- 4:20AI bot contribution boundaries
- 6:31Responsibility model for AI under CLA
- 13:43Trust and feasibility of AI contributions
- 17:19Evaluating AI-assisted contributions
- 20:30Autonomous AI bots and trust
- 22:19Next steps for AI policy in Kubernetes
related talks
01h 01m"Looks Good to Me": A Practical Guide to Handling AI-Generated Code
38mThe Cognitive Cost of AI-Assisted Open Source
24mBots & Humans: Maintaining AI agent and automation in an AI-bot web
35mAI ethics for tech writers
46mPanel Discussion: AI's Impact on Developers
01h 28mFuture of Software Programming with AI Bots at Hand
37mOur Brains in the AI Era
31mCan AI Help Bitcoin? Responsible Practices for Open-Source Development