That decision wasn't about copyrighted photos. It was about Google creating a books search index, which was allowed as fair use - just like their scanning of books for previews is. That's an entirely different situation than if Google had trained an AI to write books for sale, that contained snippets or passages from the digitized books.

The latter certainly would not be considered fair use under the reasoning given by the judge in the case. He found that the search algorithm maintained:

> consideration for the rights of authors and other creative individuals, and without adversely impacting the rights of copyright holders

and that its incorporation into the Google Books system works to increase the sales of the copyrighted books by the authors. None of this can be said about Microsoft's product. It would seem to clearly fail the tests for fair use.