Principle Failure

Is Google still succeeding at not being evil with the introduction of its new streamlined privacy policy?

Image: By dmixo6 @Flickr [CC BY-NC-ND 2.0 licence]

The right to access, correct, and delete personal information held about you and the right to bar data collected for one purpose from being reused for another are basic principles of the data protection laws that have been the norm in Europe since the EU adopted the Privacy Directive in 1995. This is the Privacy Directive that is currently being updated; the European Commission's proposals seem, inevitably, to please no one. Businesses are already complaining compliance will be unworkable or too expensive (hey, fines of up to 2 percent of global income!). I'm not sure consumers should be all that happy either; I'd rather have the right to be anonymous than to be forgotten (which I believe will prove technically unworkable), and the jurisdiction for legal disputes with a company to be set to my country rather than theirs. Much debate lies ahead.

In the meantime, the importance of the data protection laws has been enhanced by Google's announcement this week that it will revise and consolidate the more than 60 privacy policies covering its various services "to create one beautifully simple and intuitive experience across Google". It will, the press release continues, be "Tailored for you". Not the privacy policy, of course, which is a one-size-fits-all piece of corporate lawyer ass-covering, but the services you use, which, after the fragmented data Google holds about you has been pooled into one giant liquid metal Terminator, will be transformed into so-much-more personal helpfulness. Which would sound better if 2011 hadn't seen loud warnings about the danger that personalization will disappear stuff we really need to know: see Eli Pariser's filter bubble and Jeff Chester's worries about the future of democracy.

Google is right that streamlining and consolidating its myriad privacy policies is a user-friendly thing to do. Yes, let's have a single policy we can read once and understand. We hate reading even one privacy policy, let alone 60 of them.

But the furore isn't about that, it's about the single pool of data. People do not use Google Docs in order to improve their search results; they don't put up Google+ pages and join circles in order to improve the targeting of ads on YouTube. This is everything privacy advocates worried about when Gmail was launched.

Australian privacy campaigner Roger Clarke's discussion document sets out the principles that the decision violates: no consultation, retroactive application; no opt out.

Are we evil yet?

In his 2011 book, In the Plex, Steven Levy traces the beginnings of a shift in Google's views on how and when it implements advertising to the company's controversial purchase of the DoubleClick advertising network, which relied on cookies and tracking to create targeted ads based on Net users' browsing history. This $3.1 billion purchase was huge enough to set off anti-trust alarms. Rightly so. Levy writes, "...sometime after the process began, people at the company realized that they were going to wind up with the Internet-tracking equivalent of the Hope Diamond: an omniscient cookie that no other company could match." Between DoubleClick's dominance in display advertising on large, commercial Web sites and Google AdSense's presence on millions of smaller sites, the company could track pretty much all Web users. "No law prevented it from combining all that information into one file," Levy writes, adding that Google imposed limits, in that it didn't use blog postings, email, or search behavior in building those cookies.

Levy notes that Google spends a lot of time thinking about privacy, but quotes founder Larry Page as saying that the particular issues the public chooses to get upset about seem randomly chosen, the reaction determined most often by the first published headline about a particular product. This could well be true - or it may also be a sign that Page and Brin, like Facebook's Mark Zuckberg and some other Silicon Valley technology company leaders, are simply out of step with the public. Maybe the reactions only seem random because Page and Brin can't identify the underlying principles.

In blending its services, the issue isn't solely privacy, but also the long-simmering complaint that Google is increasingly favoring its own services in its search results - which would be a clear anti-trust violation. There, the traditional principle is that dominance in one market (search engines) should not be leveraged to achieve dominance in another (social networking, video watching, cloud services, email).

SearchEngineLand has a great analysis of why Google's Search Plus is such a departure for the company and what it could have done had it chosen to be consistent with its historical approach to search results. Building on the "Don't Be Evil" tool built by Twitter, Facebook, and MySpace, among others, SEL demonstrates the gaps that result from Google's choices here, and also how the company could have vastly improved its service to its search customers.

What really strikes me in all this is that the answer to both the EU issues and the Google problem may be the same: the personal data store that William Heath has been proposing for three years. Data portability and interoperability, check; user control, check. But that is as far from the Web 2.0 business model as file-sharing is from that of the entertainment industry.

Wendy M. Grossman's Web site has an extensive archive of her books, articles, and music, and an archive of all the earlier columns in this series.