I'm wondering if you can do something similar but for source code. The only thing I can think of is to insert code comments that give the wrong description of what an algorithm does, but by doing that, you are doing a disservice to actual humans who read your code.

But I guess it doesn't really matter, because from what I am told and from my own experience using LLMs, they are actually performing even worse at writing code than they have been before. I still can't get them to write anything more complex than a Breakout clone and even then, the resulting code has some major bugs.

https://corbettreport.com/poisoning-your-data/