Soul searching
In January, the Justice Department subpoenaed the search logs of Google, AOL, MSN and Yahoo. The subpoena asked them "to turn over every query typed into its search engine over the course of one week without providing identifying information about the people who conducted the searches."
(try 317966, 5776700, 98280)
View by keywords
(try "penis enlargement", "disney porn", "how to rob a bank")Three of the four complied; only Google refused.
The Justice Department claimed it needed this information in its Quixotic campaign against Internet pornography. After having a couple laws overturned by the courts on First Amendment grounds, they are apparently taking a more sophisticated approach this time around.
Search engine companies turning over large amounts of data to the federal government probably makes most of us feel a bit uncomfortable. But we can at least be comforted that the information is anonymous, right? No one can tie our searches back to us.
That idea took a big blow last month when AOL posted the March to May search logs for 657,000 anonymized users on their site, for research purposes. After not too long, someone at AOL realized this might not be the best idea in the world and took them down. By then it was too late, though; the information had been forever released to the Web.
Go to this search page and search for user 6281712. Apparently, a member of the field hockey team is thinking about giving away her special gift. Meanwhile, check out the March 20 searches by 393765 compared to those the next day.
The logs are absolutely fascinating as a glimpse into other people's lives. The Editors call 711391's logs "the greatest character study in literary history," and they might have a point. CNET gives some other interesting examples.
But as fascinating as they are, they're still anonymous. We may find it amusing that 1952262 is awfully interested in herpes, but their identity is safely protected. Right?
The New York Times decided to test that out. Could they track down user 4417749 simply from his or her search queries?
It's nerve-wracking enough to imagine Time Warner or Microsoft having access to this information. But the government? Should they act because 17556639 might be planning to kill someone? What about 1227287, who is apparently into marijuana, bomb-making and the FBI. Several commenters at Paradigm Shift compared this to the movie Minority Report, where people are arrested for crimes they have not yet committed based on a psychic vision that they will do it in the future.
Is Googling "how to kill a wife" premeditation, or just curiosity? Or research for a murder novel? Or song lyrics?
The government isn't just interested in anonymous search logs, though. Cross-referencing those with databases and designing data-mining programs powerful enough to match up the information would take an awful lot of work. Instead, the government wants to just require ISPs to hand over everything they want.
The greatest threat to our future privacy isn't warrantless wiretapping, or overzealous police, or even traffic cameras (which are even worse than parking tickets). It's simple computing power. As data storage gets cheaper and cheaper (anyone run out of space in your e-mail account recently?), eventually every single action you take online will be stored in a database somewhere. And somewhere else will be stored every item you purchased at the grocery store, every credit card transaction you've made, every plane fight you've ever been on.
All that's missing is the ability to put it all together. Guess who has the most computing power in the world right now, combined with the most advanced data-sorting programs known to man?
I'm glad Google's motto is "Don't be evil." Let's hope they keep it up.
Check out the searches of 657,000 AOL users.
View by user id(try 317966, 5776700, 98280)
View by keywords
(try "penis enlargement", "disney porn", "how to rob a bank")
The Justice Department claimed it needed this information in its Quixotic campaign against Internet pornography. After having a couple laws overturned by the courts on First Amendment grounds, they are apparently taking a more sophisticated approach this time around.
Search engine companies turning over large amounts of data to the federal government probably makes most of us feel a bit uncomfortable. But we can at least be comforted that the information is anonymous, right? No one can tie our searches back to us.
That idea took a big blow last month when AOL posted the March to May search logs for 657,000 anonymized users on their site, for research purposes. After not too long, someone at AOL realized this might not be the best idea in the world and took them down. By then it was too late, though; the information had been forever released to the Web.
Go to this search page and search for user 6281712. Apparently, a member of the field hockey team is thinking about giving away her special gift. Meanwhile, check out the March 20 searches by 393765 compared to those the next day.
The logs are absolutely fascinating as a glimpse into other people's lives. The Editors call 711391's logs "the greatest character study in literary history," and they might have a point. CNET gives some other interesting examples.
But as fascinating as they are, they're still anonymous. We may find it amusing that 1952262 is awfully interested in herpes, but their identity is safely protected. Right?
The New York Times decided to test that out. Could they track down user 4417749 simply from his or her search queries?
And search by search, click by click, the identity of AOL user No. 4417749 became easier to discern. There are queries for "landscapers in Lilburn, Ga," several people with the last name Arnold and "homes sold in shadow lake subdivision gwinnett county georgia."It took reporters from the Times to match up Arnold with her queries. But what if you cross-referenced this data with a database containing every conceivable piece of demographic information about you? Given a powerful enough computer, it probably wouldn't be too hard to assign a name to the searches, or at least narrow down the choices.
It did not take much investigating to follow that data trail to Thelma Arnold, a 62-year-old widow who lives in Lilburn, Ga., frequently researches her friends' medical ailments and loves her three dogs. "Those are my searches," she said, after a reporter read part of the list to her.
It's nerve-wracking enough to imagine Time Warner or Microsoft having access to this information. But the government? Should they act because 17556639 might be planning to kill someone? What about 1227287, who is apparently into marijuana, bomb-making and the FBI. Several commenters at Paradigm Shift compared this to the movie Minority Report, where people are arrested for crimes they have not yet committed based on a psychic vision that they will do it in the future.
Is Googling "how to kill a wife" premeditation, or just curiosity? Or research for a murder novel? Or song lyrics?
The government isn't just interested in anonymous search logs, though. Cross-referencing those with databases and designing data-mining programs powerful enough to match up the information would take an awful lot of work. Instead, the government wants to just require ISPs to hand over everything they want.
The greatest threat to our future privacy isn't warrantless wiretapping, or overzealous police, or even traffic cameras (which are even worse than parking tickets). It's simple computing power. As data storage gets cheaper and cheaper (anyone run out of space in your e-mail account recently?), eventually every single action you take online will be stored in a database somewhere. And somewhere else will be stored every item you purchased at the grocery store, every credit card transaction you've made, every plane fight you've ever been on.
All that's missing is the ability to put it all together. Guess who has the most computing power in the world right now, combined with the most advanced data-sorting programs known to man?
I'm glad Google's motto is "Don't be evil." Let's hope they keep it up.
The future ain't what it used to be.