The court granted summary judgment in favor of Google on four independent bases:
Serving a webpage from the Google Cache does not constitute direct infringement, because it results from automated, non-volitional activity by Google servers (Field did not allege infringement on the basis of the making of the initial copy by the Googlebot);
Field’s conduct (failure to set a “no archive” metatag; posting “allow all” robot.txt header) indicated that he impliedly licensed search engines to archive his web page;
The Google Cache is a fair use; and
The Google Cache qualifies for the DMCA’s 512(b) caching “safe harbor” for online service providers.
So Google is found to not violate copyright law, even though it automatically scans and indexes works online without permission. Google provides several mechansisms to de-list from their database and they plan to have a similar mechanism for the book seach project. It is simply economically infeasible to ask for permission in a system where there is no efficient way to find out who to ask. This goes for books which are under copyright but out of print, as well as web pages made publicly available on the world wide web.
The way for book search is not entirely clear, as pointed out by Lawrence Lessig, due to the reliance on an implied licence which presumably stems from the information collected by the Google web search engine having been publicly and freely posted in the first place.