Your new word for the day is 'copyfraud'. Here's the definition from Wikipedia: "Copyfraud refers to false copyright claims by individuals or institutions with respect to content that is in the public domain. Such claims are wrongful because material that is not copyrighted is free for all to use, modify and reproduce." In the current case, copyfraud also applies to materials that are license CC-by. As Peter Murray-Rust writes in the GOAL mailing list, "Springer took all the images published in its journals and stamped COPYRIGHT SPRINGER over all of them and offered them for sale at 60 USD. This included all my publications in BioMedCentral, a CC-BY Open Access journal..." In another post he notes that Oxford University Press is "charging large prices for re-use of CC-BY articles (e.g. 400 USD for use in an academic course pack for 100 students."
Let's be clear, though. Far from being 'fraud', these actions on the part of Springer and OUP are not illegal. Even if you pay OUP publishing fees to license your paper as CC-by, OUP can turn around and charge $400 for it because this is allowed by the license. Currently publishers are saying these practices are "mistakes" (and Springer, for example, has removed the images). But how long before these 'mistakes' are 'policy'? And of course, "there is the additional ongoing problem when articles which authors have paid to have Open, are hidden behind a paywall." If only somebody could have predicted that CC-by licenses would be used this way!