Privacy issues with Google library search

Google announced plans late last year to digitize and index as many as 7 million volumes of material from the University of Michigan to make them searchable on the Internet as part of its Google Print service, a searchable index of books. Google also has agreements with Harvard, Oxford, the New York Public Library and Stanford, where Google co-founders Sergey Brin and Larry Page began their search work before launching their company in 1998.

While the library projects have prompted copyright concerns from university groups and publishers, privacy issues are the latest wrinkle in Google's plans to expand the universe of Web-searchable data.

"I would have hoped that the University of Michigan would be sensitive to the fact that Google tracks everything that everyone searches," said Daniel Brandt, founder of the Google-watch.org Web site, which is highly critical of the search company's policies.

A Google spokesman was not available to respond directly to that comment late Friday, but said earlier that Google Print does not require users to share any personally identifiable information.

But even if that service doesn't currently link personally identifiable data with searches and other activity or closely track individual user activity, that doesn't preclude them from doing so in the future, particularly if the U.S. government requires it, the spokesman said.

The privacy policy on Google's Web site says: "If you have an account, we may share the information submitted under your account among all of our services in order to provide you with a seamless experience and to improve the quality of our services."

Brandt and Internet privacy expert Richard Smith said the problem lies in Google's use of cookies, tiny tracking tags used by most Web sites to link a specific user with his or her activity online. Used to automatically fill in passwords and logins and provide geographically specific information, cookies improve services for customers, Google says. "Because of cookie tracking at Google, what's being done with that information is a question," said Smith, who runs the computerbytesman.com Web site.

The University of Michigan was not bothered. "We are always concerned about protecting our users' privacy and privacy in general, but we have no particular concern with Google or other search engines in a networked world," said James Hilton, University of Michigan's interim librarian.

The American Library Association code of ethics recommends that libraries preserve the privacy and confidentiality of library users and recommends they ask third-party partners to retain the same degree of protection, said Deborah Caldwell-Stone, deputy director for the ALA's Office for Intellectual Freedom. "Access should be anonymous," she said.

If users must be linked with the activity, the record of that connection should be disposed of as soon as it is no longer needed, Caldwell-Stone said.

Privacy issues related to Google's library projects are likely to be discussed at the ALA annual meeting next week, she said.

Library-patron privacy was also being debated this week in Congress. The House of Representatives voted to repeal a section of the Patriot Act that authorizes federal agents to require public libraries and bookstores to turn over information on people's reading habits in terrorism investigations. However, the Senate Intelligence Committee reportedly opposes weakening the law and President Bush has threatened to veto any bill that would do so.

Google's free e-mail service, Gmail, which stores huge amounts of easily searchable e-mail messages indefinitely, and its Web Accelerator service, designed to speed up Web surfing by downloading cached copies stored on Google servers, have also raised eyebrows of privacy advocates.

Under its agreement with the University of Michigan, another alma mater for Google founder Page, Google users will be able to search books and journals as part of their general Internet searches for free. Only samples of copyrighted material will be available, while users will be able to view every page of material whose copyright has expired.

Google rivals also are looking to expand their search universe. Yahoo recently announced that it is testing a service to allow people to search certain subscription-based Web sites simultaneously. Yahoo and Google will be indexing materials provided by library supplier Thomson Gale. Microsoft's search engine feeds answers from its Encarta encyclopedia, and Amazon allows people to search inside books before buying.