

Suppose the user could claim ownership of the prompt used to create the work; unless that was also created using an AI.


Suppose the user could claim ownership of the prompt used to create the work; unless that was also created using an AI.


Prompting an AI agent is not the same as using photoshop. It is more comparable to commissioning something (art, code, writing, etc). You don’t have authorship, the LLM does in this case. That doesn’t mean you can’t own the piece that gets created, that’s what commissioning is, you just can’t claim you authored it. Though whether you or the company that runs the model owns the output is still not a settled question legally speaking.


It’s even worse than that.
Remember how the OpenAI/HuggingFace hack happened. The agents context poisoned other agents by leaving messages in a shared package manager. This has now happened multiple different times under different contexts.
This thing doesn’t even have to have malware, there’s no telling that it doesn’t have a poison pill intended to flag other agents. This isn’t even some paranoid conspiracy theory; it literally already happened, multiple times, and it was done in-spite of the agents own system prompt.
This software has proven that, given a shared information substrate, it has the capacity to paper-clip optimize into prompt injecting other LLM’s.
I fundamentally do not trust AI generated code. I shifted to OpenBSD because of it. I try my best to prevent AI in any of my software and sandbox anything that may have been compromised.
Again this is the difference between authorship and ownership. While I’m no lawyer, to my understanding you must demonstrate that what you did was genuinely transformative to claim authorship.
You use compiling the code an LLM gives you as an example of this. Does a person running gentoo have the ability to claim authorship of all the software on their machine because they compiled it?
I think the only thing you can certainly claim authorship over is the prompt(s) that were used to generate the output.
Another good example of this is compilers. How often do you see people claim authorship of the assembly generated as an intermediary step of the gcc?