Yes. This happens because the training data contains countless SotA "to-do" apps. This argument does not scale well to other types of software.
Most software written today (or 10 years ago, or 50 years ago) is not particularly unique. And even in that software that is unusual you usually find a lot of run-of-the-mill code for the more mundane aspects
Isn’t most standard software these days a permutation of things already done before?