Last week, AOL released 20 million web queries from 650,000 AOL users. A few hours later, they realized this was probably one of the dumbest things they could have done, quickly removed the data, and apologized. A mirror site containing the 439 MB compressed download is still available (it expands to 2GBs) and is some fascinating stuff for anyone interested in search behavior, data mining, or online marketing in general.

For those of you who don’t want to download the info, there’s a number of web interfaces up at AOLSearchDatabase.com, AOLSearchLogs.com, and AOLStalker.com to allow anyone to search the database - or you can get some of the highlights from CNN.com.

The gist is that AOL released 20 million records of real queries to their database, which happens to be powered by Google. AOL usernames were replaced with a unique number, but that number was associated with every search they did - including people, social security numbers, and addresses. As you might imagine, this caused lots of concern over privacy of data and what should and should not be allowed. Weblogs, Inc CEO turned AOL employee, Jason Calacanis even called for search engines to not keep search logs. Google apparently was not swayed to stop keeping logs, though they assure us their employees won’t release the data.

And by Wednesday, the first searcher was identified by the NY Times.

Ms. Arnold, who agreed to discuss her searches with a reporter, said she was shocked to hear that AOL had saved and published three months’ worth of them. “My goodness, it’s my whole personal life,” she said. “I had no idea somebody was looking over my shoulder.”

Looking through the data released, it’s not difficult to see why the government wants this information. Earlier this year, Google was the only search engine to refuse to hand over search logs to the Department of Justice. AOL, MSN, and Yahoo all handed over their logs when asked.

The interesting thing is that this incident may spark new legislation from Congress. Representative Edward Markey (D-Mass.) proposed a privacy bill back in Feb 2006 and is now encouraging his colleagues to take action

Markey’s bill is H.R. 4731, the Eliminate Warehousing of Consumer Internet Data Act (EWOCID). The bill would require Internet companies to destroy obsolete electronic data, and particularly data that could be used to individually identify consumers. The bill would also instruct the Federal Trade Commission to set up standards for the maintenance and destruction of data, and enforce the provisions of the law.