The Internet search-engine company Google was correct to resist a Department of Justice subpoena requesting a broad range of material from its databases, specifically 1 million random Web addresses and the records of all Google searches during a one-week period.
The Justice Department originally asked for all of Google’s Web addresses and a month’s worth of searches, then backed off when informed that could be billions of entries. This still looks like a fishing-expedition request that would not give the government what it thought it wanted.
Too bad Yahoo, MSN and AOL, which reportedly received similar requests, did not resist as well.
The government made this request of Google not in conjunction with a criminal inquiry, but to gather data to help it revive a law, the Child Online Protection Act.
That act would have required adults to use online access codes or registration systems before being able to see pornography or other material “harmful to minors.”
Two years ago the U.S. Supreme Court blocked its enforcement, noting that it was probably unconstitutional censorship and that private filtering methods were as effective as the online protection act mandate.
The government is now trying to prove the Child Online Protection Act system is better than competitive private filtering systems and that it wanted this information to help make that case, but did not ask for data that would have identified users, respecting privacy concerns.
The first problem is that proving one method of trying to protect children from online porn is more effective than another wouldn’t make that method constitutional.
Danny Sullivan has run SearchEngineWatch.com, a Web site specializing in such issues, for 10 years. He said the information the government requested would have been almost useless for this purpose. Without identifiable users, it would have been almost random data that would have given little or no insight into the likelihood of being pushed onto a porn site while searching for something else.
Google was right to resist.