Most material they're scraping is copyrighted, and reproducing it without permissions would be prohibited save for the various "fair use" provisions. Google wo

Most material they're scraping is copyrighted, and reproducing it without permissions would be prohibited save for the various "fair use" provisions. Google would probably make a "de minimis" defense of the practice. A previous case (Kelly v. Arriba Soft Corporation 2002) allowed thumbnails in search engines as fair use, but that's not quite the same. And current interpretation seems to be pretty restrictive on what's insignificant enough (witness various music sampling cases). And while any individual result may be small, doing so on an industrial scale is probably not "a trifle" and harm can probably be demonstrated. Ultimately, it would have to be tried to tell. They might also argue implied license. In cases like meta tags, that's probably reasonable. For straight-up scraping, it's more questionable.