If you want to look up an image, go to tineye.com and paste in the URL, then explore the results. (Google and I think Yandex both also have reverse image search.) The older an image is, the less likely it is to be AI. Similarly, at least at the moment, there aren't any AI images that I know of in the major repositories like Alamy. Then if you follow the links off the reverse image search page, you can usually find something that's been written about it; someplace like Alamy will give (as best they know) the caption, year, photographer, location, etc, of an image.
Why are so many showing up that you've never seen? There are hundreds of millions of paintings and likely tens of billions of photos - how could you possibly have seen and remembered every one of them? There's also more museums and collections that have put part of their works online, even before the pandemic; and some places put additional works online during the pandemic as a way to continue or increase patron engagement during the pandemic. Some places have also reached a slight critical mass of data when combining in the photos of previous exhibitions and such that they've put online.
Finally, there's been the whole digitizing-books effort over the past decade or so. We usually think of just the text that's come online, but they include millions of images as well, many of which might have seen limited release previously.