Ask HN: Do LLM code assistants rip off open source code?

I'm working on a new open source project. I want to use the GPL3 (and indeed I think I must use the GPL, because I'm incorporating algorithms that I'm learning from other GPL projects). I have this suspicion that as soon as the code is public, it will immediately be used as training data for LLM code assistants which will then regurgitate the code without any concern for the original license. What can I do about this?