Google Defies DOJ Demand for Search Records

The Department of Justice sought a court order January 18 from the U.S. District Court of Northern California demanding that Google comply with its August 2005 subpoena for millions of randomly selected online search records held by the Silicon Valley–based search-engine firm. “We intend to resist their motion vigorously,” Google Associate General Counsel Nicole Wong told the January 19 San Jose Mercury News, characterizing DOJ’s demand as “overreaching.”

The court filing indicates that the Justice Department is seeking a random sample of one million URLs findable in various Google databases, as well as “the text of each search string entered onto Google’s search engine over a one-week period (absent any information identifying the person who entered such query).” Although many libraries routinely purge the cache of their patrons’ online sessions, Google maintains server logs of all the searches its software conducts, according to the firm’s privacy policy.

The filing also revealed that the DOJ had made similar records requests of three other search engines. MSN, Yahoo, and AOL issued separate statements January 19 that they have complied to varying degrees.

DOJ officials assert they need the search records to develop a defense of the constitutionality of the Child Online Protection Act as instructed by the Supreme Court in 2004. The COPA trial, ACLU v. Gonzales, is scheduled to begin this summer before the U.S. Court of Appeals for the Third Circuit in Philadelphia.

Passed in 1998 but never enforced, COPA requires commercial websites to obtain proof of age from visitors before allowing them to view material considered harmful to minors. DOJ contends that the requested search histories would provide “a factual record in support of [the government’s] contention that COPA is more effective than filtering software in protecting minors from exposure to harmful materials on the internet.” Specifically, DOJ consultant and University of California at Berkeley statistics professor Philip B. Stark argued, reviewing users’ search terms and retrieved URLs “will help . . . measure the effectiveness of content filters in screening HTM materials,” although the filing does not explain how Stark would determine whether searches stripped of identifiable information were conducted on filtered machines.

Reacting to the DOJ requests, Rep. Edward Markey (D-Mass.) announced January 20 that he would introduce a bill requiring search-engine companies to purge personally identifiable information about their users after a reasonable period of time. “Internet search engines provide an extraordinary service, but the preservation of that service does not rely on a bottomless, timeless database that can do great damage despite good intentions,” he told Reuters.

Posted January 20, 2006.