This still seems shaky to me: wouldn't then, e.g. the Google image search pipeline perform an enormous number of copyright violations every day? They download images from a huge number of different rights holders, store them on google servers, recode them to produce thumbnails and distribute the recoded images via the image search UI. By your logic, they'd have to ask every single rights builder for permission to create derived works before they can include their images in image search.
Additionally, every time I download something from the web and store it on my hard drive, my OS will enrich it with metadata: file name and path, timestamps, local user, permissions, etc. So unless a site author grants all visitors the permission to create derived works, I'd be infringing just by visiting their web site.
Finally, if you assume that only some metadata counts as creating a derived work, Getty would have to prove that all users they threatened did in fact use Getty's "enhanced" version of the image and not the original. I find it very unlikely that the photographer herself would have used Getty's version and not herselfes, so I don't think they did that.
Additionally, every time I download something from the web and store it on my hard drive, my OS will enrich it with metadata: file name and path, timestamps, local user, permissions, etc. So unless a site author grants all visitors the permission to create derived works, I'd be infringing just by visiting their web site.
Finally, if you assume that only some metadata counts as creating a derived work, Getty would have to prove that all users they threatened did in fact use Getty's "enhanced" version of the image and not the original. I find it very unlikely that the photographer herself would have used Getty's version and not herselfes, so I don't think they did that.