Hello, astrophotographer here.

The unique thing about astrophotography is that the subject that we are imaging is (relatively speaking) static. Any slight deviation, even the smallest detail, that is not structurally accurate is PAINFULLY obvious.

As an aside, we do use traditional neural networks for certain processing steps. Deconvolution and noise reduction are very common. However, these are only helping in correcting errors caused by viewing conditions. They are not the "imaginative" AI that you are dreading.

I had a feeling something like this was true, thank you for confirming it!

I would actually be more worried about a knowledgeable human engaging in clever Photoshop alterations, or pasting different images together. Is this an issue in astrophotography? Are there ways to easily detect this?

There is a category called "Annie Maunder Open Category", which allows for creative use of generative AI, as long as it is declared, otherwise it is banned. You have to take that slightly with a pinch of salt though, as many of the images don't represent what could be seen with the naked eye. It could be as simple as a wide spectrum sensor (which I don't think anyone could claim is AI) through to alignment of many 100s of images (for which the algorithms in use may have been considered AI a decade or 2 ago).