Michael Zimmer.org

Microsoft has fired a new salvo into the search privacy wars, announcing it will delete IP addresses from the Bing search engine logs after 6 months.

Microsoft has decided to take the lead in search privacy and agree to the European Union’s demand that data retention be cut to six months. Previously, Microsoft de-identified its search logs immediately, but didn’t purge the IP address until 18 months. Now, de-identification still takes place immediately, and the IP addresses are completely removed in 6 months.

Yesterday I posted that Cuil, the supposed “Google-killer” search engine that once took pride in not keeping any logs of its users’ activities, had dramatically altered its privacy policy, effectively stripping it of the strong privacy-protecting language it originally contained. Since then, I’ve received 3 communications from Cuil.
The first was a tweet promising an email (not yet received) as well as the assertion that “For the record, we still don’t keep logs or store personal info”. It is kind of them to tell me this via Twitter, but unless their …

Remember Cuil, the search engine launched in 2008 that was supposed to be a Google-killer? Didn’t think so.
Anyway, one of Cuil’s touted competitive advantages was that it didn’t track user search queries. Its original privacy policy (dated July 27, 2008) went to great lengths to make users feel comfortable about the privacy of their search activities, opening with this impressive declaration:
Privacy is a hot topic these days, and we want you to feel totally comfortable using our service, so our privacy policy is very simple: when you search with Cuil, …

For quite some time now, I’ve been writing about how “search” has become the center of gravity of our informational ecosystem, and that a primary externality of our dependence on search has been the threat to privacy. On numerous occasions I’ve called on Google to engage in value-conscious design in order to protect user privacy, and specifically argued for the creation of a Google Data Privacy center where users can see exactly what data Google has collected about them from their expansive infrastructure of dataveillance, edit or remove this data …

Some filmmakers have produced an amazing series of episodes based on one person’s searches discovered in the AOL search data release debacle. Here’s the trailer:

And the description from the website where you can view them all:
August 4, 2006, the personal search queries of 650,000 AOL (America Online) users accidentally ended up on the Internet, for all to see. These search queries were entered in AOL’s search engine over a three-month period. After three days AOL realized their blunder and removed the data from their site, but the sensitive private data …

Well, according to Microsoft.
As part of the on-going battle among major search engines related to search data retention policies, Microsoft has published this chart attempting to summarize and compare the state of anonymization in the search industry (click to enlarge).

Few techncial details about how the companies enforce these policies have been made public, and as Joris van Hoboken reminds us, de-identification is not the the same as anonymization. There’s good work to be done to confirm, validate, and test the procedures claimed by each of the major search engines.

There has been a lot of movemnet recently regarding how search engines treat the vast amounts of user data they collect.
In early 2007, Google announced it would “anonymize” its user search logs after 18-24 months. Later that year, Google reluctantly decided to add an expiration date to its web cookie, while Ask.com (unsuccessfully) tried to gain market share by giving users almost complete control over whether any data is collected. Earlier this year, under pressure from EU regulators, Google announced it would anonymize its search logs after 9 months. And …

Quickly, as I’m rushing out the door to teach: Google has announced it will (reluctantly) shorten the amount of time it holds fully-identifiable information, and will “anonymize” search records after 9 months, rather than the current 18 months.
This is big news…more later….

Some former Googlers have launched a rival search engine named for the Gaelic word for knolwedge, Cuil.
Cuil (pronounced like “cool”), which claims to have an index three times the size as Google and ten times as Microsoft, aims to provide a difference kind of search experience than its friends in Mountain View:
Rather than rely on superficial popularity metrics, Cuil searches for and ranks pages based on their content and relevance. When we find a page with your keywords, we stay on that page and analyze the rest of its content, …

After coming under attack for refusing to add a simple hyperlink to help users find their privacy policy, Google has added the word “privacy”, with a link to its privacy policy, to its home page (image via Google):

Google, for whatever reason, isn’t directly acknowledging that there was public pressure to take this simple step. The announcement on Google’s public policy blog frames it in a self-congratulatory manner, noting how it’s putting “users’ privacy first and foremost” and strengthening user trust, etc, etc. And their main blog relates some odd story …

Much of Google’s resistance to adding a link to its privacy policy on its homepage seems to boil down to little more than aesthetics. As Steve Langdon, a Google spokesman, puts it: “We do believe that having very limited text on our home page.”
Certainly, Google’s simple homepage design has its value. Much of Google’s early success can be attributed to the speed at which its homepage loaded, especially with early Internet users relying on dial-up connections. Compared to Yahoo’s homepage that was loaded with images and display ads, Google’s spartan …

As we know, Google refuses to place a link to its privacy policy on either its homepage or search results pages. It has been pointed out (thanks, Chris) that the California Online Privacy Protection Act of 2003 requires the operator of a commercial Web site that collects personal information about users to “conspicuously post its privacy policy on its Web site.” When asked about this by the NY Times, Google responds, in part, that “because the privacy policy is easily found by using the search box on the home page, …

Prompted by Google’s resistance to cluttering its homepage with a link to its privacy policy, I decided to take a quick tour of the major search engines to compare the relative visibility of their privacy policies.
AOL, Microsoft, and Yahoo! have visible privacy policies on both their homepages and search results page (which is especially important if you use automatic search toolbars on browsers without visiting the homepage).
Neither Ask nor Google provide direct links to their privacy policy on their homepage or results pages.
[**UPDATE**  On June 18, 2008, Ask.com added a …

I’ve long complained about how it takes at least 2 clicks to get to Google’s privacy policy from its homepage (3 clicks if you count its new Privacy Center splash page). And that’s only if you happen to click on “About Google,” and then happen to find the “Privacy Policy” link at the bottom of that page. Nowhere on Google’s homepage, nor on its search results page, will you find the words “privacy policy.” A user must actively seek out the world’s largest search engine’s privacy policy — one would …

Yahoo has issued a press release on their new ad sales system, AMP! that provided a few new insights into this “powerful new online advertising platform.” The New York Times broke this story earlier, which I blogged about here.
Interestingly, the Yahoo release avoids using the phrase “behavioral targeting” (which is mentioned in the NYT piece), and instead note how the system will allow publishers to connect with “their exact target audiences across the increasingly fragmented Internet” and enable “precise geographic, demographic, and interest-based …