Currently the InternetArchiveBot does not run on Wikimedia Commons, as many images are uploaded from external sources, for proper attribution the copyright status of an image will have to be fetched from sites such as Verizon's Flickr and others, though currently all of this information IS copied when imported it is not impossible for Flickr-users (or other providers of media) to change the status of the copyright, or for websites to cease to exist, for that reason I would like to suggest running the InternetArchiveBot on Wikimedia Commons and allowing it to cache websites from the upload date and the date a (new) link was added (which can happen in cases of misattribution).

The InternetArchiveBot should mostly scan "Source" but also all other links in the description and other areas of a file as sometimes media files can have multiple links for attribution or older files can have their sources in the description.