But copying and reproducing snippets from wikipedia doesn't do any harm to their technology infrastructure. It would be a negligible amount of hits to their sys

But copying and reproducing snippets from wikipedia doesn't do any harm to their technology infrastructure. It would be a negligible amount of hits to their system to grab and update the cached snippets. And you are correct that the terms of service and license are different (even if there is a bit of overlap) but it's the license that matters when you are "copying" the content and reproducing it on your own, and the license is either CC BY-SA or GFDL both of which have no ability to prevent any one specific person from using the information. >In summary, they reserve the right to block access in response to abuse. This isn't abuse, its use. If copying the data from wikipedia and serving it up with attribution is abuse, then they need to update their terms of service and licenses to explicitly say so (which would pretty much end wikipedia). If Google is harming their infrastructure by causing excessive load doing the scraping, wikipedia has every right to block them. But Google is not (as far as we know), and therefore wikipedia does not have any moral right to block access, and Google has no moral or legal obligation to stop scraping and reproducing the content on their own servers. And if wikipedia does block Google from the service, Google has every right (legally and morally in my opinion) to find an alternate way of getting the data and reproducing it that doesn't access wikipedia's servers directly.