The Subplot project prefers to not accept contributions made with generative AI systems ("LLM", "AI", chatbots). While we can't prevent contributors from using such systems to explore and understand our codebase, nor to assist in making changes to propose to Subplot, we strongly advise against its use.

We reserve the right to reject any contribution we suspect has been made with generative AI technology. This includes, but is not limited to, issues, comments, and patches.

If you feel you need to use generative AI in order to communicate effectively, for example if English is not your primary language, then we would rather you submit issues and patches in your native language. We can use translation tools, or may actually speak enough of your language to communicate, and this way we can be confident that any misunderstanding will be on our part and not on the part of tools you used.

If you want to use Subplot source code, documentation, or other artifacts to train your own generative AI then please be aware that Subplot is licensed under the GPL. Any use of Subplot material to train or fine-tune an LLM constitutes the creation of a derived work as far as we are concerned and, as such, the GPL automatically applies to your LLM as well as anything it subsequently generates.

Discussion

Generative AI technology is increasingly used to make changes to free and open source software. This has many problems. This is not an exhaustive list of them, but they convince us.

We do not consider any training or use of generative AI technology to be acceptable.

• Generative AI is inherently anti-human and anti-labour.
  • Tante: AI as a Fascist Artifact
  • Vetch: Back up a layer
  • Stovall: Eugenics Powers IQ and AI
  • Dove-Viebahn: Sundance 2026: The Masculinist and Eugenicist Origins of AI Are Writ Large in Documentary ‘Ghost in the Machine’
  • Koebler: 'AI Is African Intelligence': The Workers Who Train AI Are Fighting Back
  • Behal: ‘In the end, you feel blank’: India’s female workers watching hours of abusive content to train AI
  • Palma: How Business Can Reduce Labor Cost With AI
• It's an ecological disaster.
  • Gorey: Data Drain: The Land and Water Impacts of the AI Boom
• It's built without permission to use the source material and disregards open source licenses.
• The business is suspicious and unprofitable.
  • Shonk: What Are Circular Deals?
• It's bad for its users, both individuals and organizations.
  • Saini: AWS CEO Explains 3 Reasons AI Can’t Replace Junior Devs
  • Cro: Contributor Poker and Zig's AI Ban
  • Tian, Zchang: Learners' AI dependence and critical thinking: The psychological mechanism of fatigue and the social buffering role of AI literacy
  • Koebler: Your AI Use Is Breaking My Brain
• Hides origin of code, thus exposing us to later accusations of using code we have no permission to use.
  • Jandhyala, Kim, Bhattacharayya: Copyrights, Professional Perspective - IP Issues With AI Code Generators
• Generative AI services are effectively doing a DDoS attack on websites and Git services.
  • Nagy: iocaine
  • Iaso: Anubis
  • Webb: HardenedBSD May 2026 Status Report
• The terminology used to discuss the technology is obfuscating and anthromorphizing.
  • the machines do not "learn", a statistical model is built from input data (with human guidance) to produce output that is plausible
  • the machines do not "hallucinate", they produced incorrect output
  • Marwala: The Concern Around Saying AI ‘Hallucinates’
• The technology is extremely dependent on services provided by big technology corporations, and that dependency prevents software freedom.

Even if everything else were perfect and fine and good, generative AI takes away the fun from software development, at least for us.