The Hubzilla project, a privacy-focused open source publishing and social networking platform, adopted a policy banning contributions that include content generated in part or in full by large language models, diffusion models or similar systems, covering source code, text and images in its Git repositories as well as merge requests, code reviews and comments. Contributors may still use generative AI privately to understand, debug and review the codebase or do research, but must not submit tool-generated content to the core or official repositories, and dependencies created entirely or predominantly by generative AI will not be used. Hubzilla cites maintainer workload and review burden, licensing uncertainty that could create liability, and environmental, labor and bias concerns, and says the policy is adapted from the Subplot project’s generative AI policy and the OpenJDK interim policy on generative AI.
Hubzilla adopts a no-LLM contribution policy
The Hubzilla project, a privacy-focused open source publishing and social networking platform, adopted a policy banning contributions that include content generated in part or in full by large language models, diffusion models or similar systems, covering source code, text and images in its Git repositories as well as merge requests, code reviews and comments. Contributors may still use generative AI privately to understand, debug and review the codebase or do research, but must not submit tool-generated content to the core or official repositories, and dependencies created entirely or predominantly by generative AI will not be used. Hubzilla cites maintainer workload and review burden, licensing uncertainty that could create liability, and environmental, labor and bias concerns, and says the policy is adapted from the Subplot project's generative AI policy and the OpenJDK interim policy on generative AI.
Source: Hubzilla