The book-scanning contract between Google and the University of California has made the Association of American Publishers (AAP) even more distressed over Google's project to digitize millions of volumes from libraries.

Last year, the AAP sued Google on behalf of five of its members -- The McGraw-Hill Companies, Pearson Education, Penguin Group USA, Simon & Schuster and John Wiley & Sons -- alleging massive copyright infringement in the Google Books Library Project.

But the University of California contract, released publicly two weeks ago, has the AAP even more dismayed than before, said Allan Adler, the AAP's vice president of legal and government affairs.

In particular, the University of California agreement contains eyebrow-raising provisions that don't appear in the Google Books Library Project contract with the University of Michigan, which is also available publicly, Adler said.

"Some aspects of the University of California contract that differ from the University of Michigan contract raise additional questions both about Google's liability and the possible liability of the universities," Adler said.

Adler points to the University of California contract's definition of university library patron as "all individuals and organizations that the University Libraries serve from their Web sites." The contract lets the university use its digital copies for services offered to its patrons, subject to copyright law.

Adler finds that the "patron" definition seems broad and could include the general public, not just faculty, students and staff. The contract also allows the university "to recover copying costs" incurred in providing services to patrons. "What's the extent of copying that this agreement would permit the University of California library patrons to engage in?" Adler wonders.

Jennifer Colvin, a University of California spokeswoman, said that some online library services are available to the general public and others are limited to university community members. She stressed that the university hasn't decided how it will use its digital copies, but that the uses will comply with copyright law.

Adler also noted that the contract, unlike the Michigan agreement, lets the University of California share up to 10 percent of its Google-provided digitized book collection with other libraries and educational institutions for noncommercial purposes. "There's a much tighter set of restrictions on what the University of Michigan can do with the digital copies it receives from Google," Adler said. The University of California seems to have negotiated access for its copies to a more broadly defined universe of so-called patrons, he said.

Asked about the differences between the two university contracts, a Google spokeswoman said that each agreement is individually crafted to meet each partner's requirements. She declined to explain why the sharing limit is set at 10 percent.

Adler also saw indications in the University of California contract that Google may have grown more concerned and aware of potential liabilities since late 2004, when it signed the Michigan contract.

For example, the California contract stipulates that Google will provide digital copies to the university "via a network connection." Based on the Michigan contract, the AAP assumed Google provides the digitized books to the universities in storage hardware like CDs and hard disks. By hosting the university copies on its servers, Google might be attempting to avoid distributing these copies in a physical format, Adler said. The Google and the University of California spokeswomen separately declined to say how the digital copies are being transferred.

The California agreement also forces third-party universities and institutions that gain access to the University of California digital copies to indemnify Google in writing from liability related to how these organizations use the material, Adler notes.

Last month, the University of California became the latest institution to join the Google Books Library Project, agreeing to let Google scan and digitize millions of volumes from its libraries. Weeks later, the university publicly released the contract to satisfy a "general interest" in the document, including a request by the IDG News Service.

The Authors Guild and three authors also sued Google for alleged copyright infringement related to the library book-scanning project last year. The AAP and the Authors Guild both allege Google needs to obtain permission from copyright holders before scanning in-copyright books it obtains from the libraries.

Google acknowledges that it scans the books and makes their full text searchable by users of the Google Book Search service. However, Google maintains its activities are legal because it only displays short text excerpts and bibliographic information for in-copyright books.

To this, the AAP retorts that the very act of making a digital copy of an in-copyright book without permission from the copyright holder amounts to infringement. The AAP also condemns as infringement Google's provision of a digital copy of in-copyright books to its university partners. The AAP also contends Google has no right to store on its servers the text of copyrighted books without permission, particularly because Google is using that copyrighted text to power its Book Search service.

The University of California contract is available here. The University of Michigan contract is here. Other Google partners include Harvard University, Stanford University and Oxford University. There is no money exchange or revenue sharing between Google and the universities, according to Google.