The U.S. Department of Justice will be able to get a look at 50,000 Web addresses from Google’s index, but it can’t have 5,000 random keywords used in Google searches, a federal judge ruled Friday.
Judge James Ware of the U.S. District for the Northern District of California said in his opinion that Google must hand over the Web URLs placed in its index by automatic “searchbots”. The government says it will use those addresses in building its case for reinstatement of a law shielding children from pornography, which was overturned as unconstitutional in 2004.
But Ware upheld Google’s fight to maintain the privacy of the actual words people use when searching that Web index. Google had maintained that producing search terms would impede its service by diverting resources to pull those keywords and would also impair its goodwill with users, who might use the search engine less as a result.
After a Tuesday hearing of arguments from both parties, the court had intimated that it would grant only part of the DOJ request. It was widely expected that the judge was more likely to order the URLs released.
“This is a clear victory for our users and for our company,” Google associate general counsel Nicole Wong wrote in a post on the official Google blog after the verdict. “We will always be subject to government subpoenas, but the fact that the judge sent a clear message about privacy is reassuring. What his ruling means is that neither the government nor anyone else has carte blanche when demanding data from Internet companies.”
Ware’s opinion cited privacy concerns from users entering their Social Security numbers or credit card numbers in order to check if these are available over the Internet at some site. He also pointed to the habit of “vanity searches” on Google, in which users look for their own names, sometimes in combination with other keywords. “While a user’s search query reading ‘(user name) stanford glee club’ may not raise serious privacy concerns, a user’s search for ‘(user name) third trimester abortion san jose’ may,” he wrote.
And while Google’s own terms of service specify only that personal information such as names, e-mail addresses and billing information will be protected in their records, Ware said evidence given by government experts that as many as one-quarter of all Internet searches are for pornography indicated that users had an expectation of privacy, and that turning search terms over to the government could erode their trust in Google.
“This concern, combined with the prevalence of Internet searches for sexually explicit material … gives this court pause as to whether the search queries themselves may constitute potentially sensitive information,” he wrote.
The Justice Department had progressively reduced the amount of data it said it wanted from Google. The original subpoena filed in [[??]] asked for all Web pages in the company’s index as of July 31 2005, together with all search queries for the months of June and July 2005. After negotiation with Google, that request was cut down to a random sample of 1 million URLs from the index and all search queries for a week. The final request of only 50,000 Web pages and 5,000 keywords represents a small fraction of even the re-negotiated demand.
Google argued during last week’s hearing that the information the government wanted concerning the availability of adult material on the Internet could just as easily be found using a metasearch engine such as Dogpile.
And under questioning from Ware, government attorneys admitted they could probably make their case for the prevalence of pornography on the Web without Google’s data.
