Zimmer, M. (2012). The ethical (re)design of the Google Books project. In iConference ’12 Proceedings of the 2012 iConference, 363-369. DOI: 10.1145/2132176.2132223

Today, the Google Books project is at a relative standstill — lawsuits against the project remain outstanding as the courts rejected a proposed settlement agreement. The failure of the original vision for the Google Books project to become fully realized presents us with a unique opportunity to ensure that whatever final form Google Books will take in the future, it is designed to support the values respected within the domain of information ethics. This paper will proposed an ethical re-design of the Google Books project, focusing on three core ethical values of primary interest to librarian and information professionals: privacy, intellectual freedom, and public access to information. Advocating for these values in the next iteration of the mass digitization service can help ensure that the informational norms of the library are embraced and upheld.

Section 3.7(e) Google’s Exclusion of Books

Google may, at its discretion, exclude particular Books from one or more Display Uses for editorial or non-editorial reasons. However, Google’s right to exclude Books for editorial reasons (i.e., not for quality, user experience, legal or other non-editorial reasons) is an issue of great sensitivity to Plaintiffs and Google.  Accordingly, because Plaintiffs, Google and the libraries all value the principle of freedom of expression, and agree that this principle is an important part of GBS and other Google Products and Services, Google agrees to notify the Registry of any such exclusion of a Book for editorial reasons and of any information Google has that is pertinent to the Registry’s use of such Book other than Confidential Information of Google and other than information that Google received from a third party under an obligation of confidentiality.

Google Book Search is a massive undertaking.  Its goal is to provide unprecedented access to digital copies of all kinds of literature – a vast library of material the likes of which has never before been assembled in a single resource.  However, the project has been criticized, not for what it is including, but for what it isn’t.  Because Google Book Search has the potential to be such a widely used resource, it has the responsibility to reflect an equally diverse range of opinions and perspectives in its collection.  The editorial decision to not include a book in this project could be considered censorship and sufficient justification for such exclusion should be required of Google.  Censorship of this kind could come in two forms – first in the decision to withhold a digitized book from display in the database or, second, by choosing not to digitize a book in the first place.  The first form has already been discussed in Michael Zimmer’s blog, so the second will be the focus of this piece.

The Google Book Search Amended Settlement Agreement carves out space allowing Google to exclude certain books from “Display Uses” for both editorial and non-editorial reasons.  Alexander Macgillivray, former Google employee and head of the legal team spearheading the settlement, has gone on record assuring the public that this exception merely reserves Google’s right to exclude, but that Google has “ABSOLUTELY NO PLANS to remove any books for editorial reasons” (emphasis in original).  Despite this assurance, Section 3.7(e) implies that if, for any reason, Google does exercise this right then notice will be provided to a registry maintained by the Publisher’s Guild.  What Mr. Macgillivray and Section 3.7 do not address is the ability of Google and participating libraries to editorially exclude books from the digitization process in the first place.  Specifically, Section 3.7 discusses the exclusion of books from “Display Uses” which are defined within the Amended Settlement Agreement as “Snippet Display, Front Matter Display, Access Uses and Preview Uses” (Article 1, Section 1.52 at p. 8).  Each of these terms is further defined as display and use of material after it has already been digitized.  Again, there is no mention of the possibility that either Google or participating libraries may withhold certain books from the initial scanning process for editorial reasons.  Additionally, no requirement exists for the provision of notice to the registry for such exclusion of books from digitization.

It is conceivable that books may be excluded from the scanning process for completely legitimate reasons.  Perhaps a book is too fragile or irreplaceable to risk scanning it.  Perhaps the print is too faint or pages are missing.  These are all reasonable, non-editorial justifications for omitting a book from the digitization process.  However, it is equally conceivable that a book may not be scanned because it is too controversial, too outdated, or simply too unpopular to merit digitization.  One could easily imagine a librarian or Google scanning technician setting aside a copy of Little Black Sambo or The Anarchist’s Cookbook in order to preemptively avoid controversy.  To date there is no evidence that these books have been omitted from the scanning process, the point is that such exclusion based on the mere anticipation of ruffled feathers amounts to a priori censorship.  Without a reporting requirement similar to that included in Section 3.7, there is no transparency in the process.

Although this possibility may seem somewhat alarmist and is certainly hypothetical, a close reading of the Settlement Agreement should give one pause.  For instance, the Agreement clearly allows Google to pick and choose books from a library’s collection.  Even the definition of “Collection” within the agreement implies a certain amount of cherry-picking: “’Collection’ means the Books held by a Fully Participating Library or a Cooperating Library that have been Digitized or are targeted for Digitization pursuant to a Digitization Agreement between Google and such Fully Participating Library or such Cooperating Library, which Books may be some or all of such Fully Participating Library’s or such Cooperating Library’s holdings” (Article 1, Section 1.30 at p. 6, emphasis added).  Because these Digitization Agreements are not part of the public record, it remains unclear why only “some” of a participating library’s collection is being scanned and how such decisions are made.

More insidiously, in the final analysis Google Book Search is primarily a commercial enterprise.  Therefore, it isn’t unreasonable to assume that controversial, outdated, or unpopular material may be preemptively excluded from the scanning process because the potential market for such works is small or nonexistent.  Such exclusion may also shield Google from potentially costly litigation.  The alternative argument could certainly be made that Google has historically been inclusive of unpopular speech in its search index and that, “Nearly all known instances of the removal of content from Google’s index were, in one way or another, legally required” (see Zimmer blog post).  Mr. Macgillivray’s public statements also seem to indicate that this policy will extend to Google Book Search and that “Google does not plan to omit any books from the service, just as we have not omitted any books from our scanning based on their content or copyright status”.  While this promise and general policy of inclusion are reassuring, past practice is in no way a guarantee of future behavior.  Additionally, the inclusion of a website in Google’s search index costs almost nothing.  Google’s massive expenditure to refine the scanning process and the considerable per unit expense of digitizing a book may affect the calculus of inclusion.  Again, without some sort of public reporting requirements, we have no way of knowing if books are being excluded, why they’re being excluded, and how those decisions are being made.

If Google’s mission is “to organize the world’s information and make it universally accessible and useful” and if Google Book Search aspires to that goal, then the threshold for not digitizing a book should be quite high indeed.  Google and the Amended Settlement Agreement have created what amounts to a de facto monopoly – they are the sole online provider of these books.  No one else has the resources, technology, or access to material that Google does. Therefore, the threshold for exclusion should certainly be higher than commercial considerations of a book’s potential value in the marketplace or fear of controversy.  If, for some reason, a user is not able to access the totality of material at Google’s disposal then some justification should be required.  Public notification of these justifications should also be a necessary component for transparency – anything less is simply censorship.

It is important to realize that before Like.com was helping people find shoes and watches online, its technology was the core of Riya, a photo sharing and search site that allowed users to upload, tag and search images based on facial recognition technology. Users simply uploaded their photo library to Riya, tag faces in a handful of the photos, and then Riya’s facial recognition technology took over and attempt to automatically tag different faces it “recognizes” so that you don’t have to. Riya’s vision was to expand this facial recognition beyond an individual’s library, and eventually reach every digital photo scattered across the Web.

Over 4 years ago, I wrote about the privacy implications of Riya’s facial recognition technology here and here. I warned then:

Sure, it would be great if I could easily search for pictures of me on the web, but do I want others to be able to do that as well? Could this provide a useful tool for stalkers looking for a particular person? Abusive partners trying to track down their victim? (Note that Riya also uploads and indexes all the metadata related to your photos, including the date and time is was taken, when it was uploaded, etc. Users can also tag and search photos based on location). …there are externalities once all of the images of our daily lives (and their related metadata) are uploaded to the Internet, indexed, searchable, and accessible to all.

Google almost bought Riya back in 2006, but acquired Neven Vision instead, and soon integrated basic facial recognition into its image search results, and, later, into Picasa and Orkut.

With the purchase of Like.com, Google obtains the original Riya technology. Coupled with other recent social networking related acquisitions, Google appears poised to make a run at Facebook, with Riya providing the means to automate photo tagging at a scale Riya could only have dreamed of. What remains to be seen is how carefully Google will consider the privacy implications of unleashing powerful face recognition technology on throngs of users.

Google’s explanation was that this privacy invasion was a mistake, and happened because some code inadvertently made its way into the Street View vehicles’ software. While I trust Google that this was a mistake, and that the data wasn’t used for anything, it reveals a significant lack of control over what its fleet of vehicles are doing — and what they are capable of doing without Google apparently knowing. It also reveals yet another example of how Google failed to recognize and address possible privacy issues related to the the fact they are deploying an army of vehicles to harvest information about the physical (and now wireless) terrain.

But not to be left out of the recent spate of dotcom executives making ignorant statements about online privacy, Eric Schmidt, Google’ CEO, had this response when asked about possible EU charges against Google for the WiFi privacy invasion:

“no harm, no foul”

and

“Who was harmed? Name the person.”

and

it was “highly unlikely” that any of the collected information was “useful” and that there appeared to “have been no use of that data.”

So, once again, we have the person in charge of a dominant Web company, a company in control of huge amounts of personal data about millions of users, defining privacy concerns solely in terms of the potential (or real) harm that could occur.

Schmidt’s harm-based conception of privacy supposes that so long as the data can be protected/prevented from being used to cause harm, the privacy of the subjects is maintained. Since no one was hurt, Schmidt appears to say, then what’s the big fuss?

Such a position ignores the broader dignity-based theory of privacy. This view recognizes that one does not need to have a tangible harm take place in order for there to be concerns over the privacy of one’s personal information. Rather, merely having one’s  information stripped from the intended sphere (personal WiFi network), and amassed by passing vehicles operated by the world’s largest search engine becomes an affront to the subjects’ human dignity and their ability to control the flow of their personal information.

But Schmidt doesn’t see things this way. Why? My theory is because he’s an engineer, not an ethicist. To an engineer, if the data has no obvious use-value, and no one was hurt, then all is good with the world. To an ethicist, harvesting personal information from the spheres of one’s personal WiFi network (whether the network was open or closed) is a privacy violation.

Until the computer scientists and engineers running the companies that possess so much of our personal information start to understand online privacy from contextual and dignity based frameworks, our privacy remains in peril.

[image source]

In January of this this year, Google was the victim of a cyber-attack that originated from China. In its description of the attacks, Google noted that the ultimate target of these attacks included human rights activists, with the apparent intent to surveill their communications on various Google platforms. This, along with what Google described as “attempts over the past year to further limit free speech on the web”, prompted Google to announce that it was “no longer willing to continue censoring our results on Google.cn”.

This appeared to a major breakthrough, as numerous attempts to get Google to end their censorship of the Web and to fully embrace a commitment to protecting human rights had all previously failed.

Over three months have passed since this announcement, and today we finally learned Google’s “new approach” to China: pack up and move to Hong Kong.

Starting today, users visiting Google.cn will be redirected to Google.com.hk, Google’s Hong Kong search portal, where search results will be provided free from the filtering Google had previously been performing on Google.cn. Google is touting this as ending censorship in China, but, as Siva Vaidhyanathan has pointed out, that really isn’t the case. It’s an end-around. A slight-of-hand.

Google is simply routing customers to the Hong Kong site, where it — presumably — doesn’t have to abide by China’s laws regarding censorship (to get a .cn domain, you have to agree to comply with all applicable laws, regulations and policies of the China’s governmental agencies and the China Internet Network Information Center). So, while Chinese users will now be able to access unfiltered results at Google.com.hk, they could always do that simply by going to Google.com. The Chinese government can now simply block access to individual websites through its firewall like before. We’re back to square one.

There have been suggestions that China must now decide whether to censor a Hong Kong-based service, something it has been hesitant to do. I don’t see it this way. China doesn’t need to touch the Google.com.hk site; all it needs to do is continue to block access to the sites it doesn’t want people from Chinese IP addresses to view. It matters little whether the link was clicked from a search results page on Google.com or Google.com.hk or even Baidu.com. China won’t need to tread onto Hong Kong’s quasi-independence to continue its censorship.

A final point targeted at Google’s apparent skirting of its ethical responsibility. By refusing to stay and fight Chinese censorship directly (such as how Wikipedia has tried to stand firm), Google is waking away from any influence it might have towards ending oppressive censorship in China. When it first announced its cooperation with Chinese censors, Google defended its actions at a hearing before the Committee on International Relations of the U.S. House of Representatives by saying this:

If Google were to stay out of China, it would remove powerful pressure on the local players in the search engine market to create ever-more-powerful tools for accessing and organizing information. Google’s withdrawal from China would cede the terrain to the local Internet portals that may not have the same commitment, or feel the competitive pressure, to innovate in the interests of their users.

Apparently Google is now content with ceding the terrain, and leaving Chinese Internet users at the mercy of “local Internet portals” like Baidu, who has been described as being “the most proactive and restrictive online censor in the search arena.”

I know Google is trying to do the right thing here, and it hopes it can deliver unfiltered results to China from Google.com.hk (or force China to take some kind of action against the Hong Kong site). But I fear this move will instead result in further failure to serve the interests of Chinese Internet users, and another lost opportunity to fight oppressive online censorship.

UPDATE: As predicted, China is now blocking some access to the redirected Google.com.hk site, as well as some of the linked sites.

[image via hunxue-er]

Use information to provide our users with valuable products and services.

Develop products that reflect strong privacy standards and practices.

Make the collection of personal information transparent.

Give users meaningful choices to protect their privacy.

Be a responsible steward of the information we hold.

The principles are further explained in a video on the Google Blog (interestingly posted by an engineer, not one of Google’s legal/policy folks).

I like these principles; they are something every organization should commit to and strive for. The problem is, Google hasn’t adhered to them quite as closely as they’d want you to believe. Let’s consider each:

1. Use information to provide our users with valuable products and services. This isn’t so much a privacy principle as it is a disclaimer for what Google purports to do with all the data it collects about its millions of users. Google tracks what we do in order to know whether our search for “Paris Hilton” is about the blond or the hotel. This principle merely presents the value proposition for Google’s potential violation of user privacy.

2. Develop products that reflect strong privacy standards and practices. A very important goal, but the product featured in Google’s video, off-the-record chats in iChat, isn’t providing the kinds of privacy protections that most consumers or advocates clamor for. Certainly, being able to control (to an extent) whether my chats are logged is a way to protect my privacy, but what about IP logging or behavioral targeting? Perhaps Google doesn’t want to bring up its current data retention policies given Microsoft’s recent announcement. And perhaps it doesn’t want to actively promote one of its truly innovative privacy protecting product — the Google Advertising Cookie Opt-Out Plugin — since the more users who install the plugin, the less valuable its advertising platform becomes.

3. Make the collection of personal information transparent. Despite what Google claims about Dashboard, there remains an enormous lack of transparency regarding the collection of user information (Google Analytics comes immediately to mind). If Google was committed to transparency, it wouldn’t have resisted placing a link to its privacy policy on the homepage. If Google was committed to transparency, its behavioral targeting system would be opt-in and would provide a conspicuous link to “Ad Privacy Preferences”.

4. Give users meaningful choices to protect their privacy. Google touts the ability to report problems in Street View and the removal of one’s search history as examples of this principle. Of course, the Street View example has a horrid history, and removing your search history only removes it from that product’s interface, not from Google’s main server logs. That’s a limited choice, not a fully meaningful one.

5. Be a responsible steward of the information we hold. I have faith that Google is indeed being responsible with our information, and that it is keeping it secure. But while security is often necessary to ensure privacy, it certainly isn’t a sufficient condition, and the gaps in the preceding principles overshadow Google’s good stewardship.

In summary, I do give Google much credit for the steps they’ve taken in recent years to improve its privacy practices and communication. But too often its rhetoric is too self-congratulatory, and fails to recognize serious gaps in its approach to user privacy.

These principles are vital, and I hope Google continues to strive to meet them. There is much work still to be done.

Remember how hard we gad to fight to convince Google to include a link to its privacy policy on the Google.com homepage?

Remember how Google argued “we do believe that having very limited text on our home page is important” and that it was pitched as some great sacrifice to include the word “privacy” and disrupt the homepage’s aesthetics?

Remember how, just a month ago, Google argued that a “clean, minimalist approach” to their homepage gives users “just what they are looking for first and foremost” — thus only the search box appeared when you visit Google.com, with all the other content (including the privacy policy link) only fading in if you happen to move the mouse?

Apparently none of this is nearly as important as shilling promoting Google’s new phone, the Nexus One. If you visit Google.com today, you’ll notice everything faded out except a prominently placed Nexus One advertisement:

Move your mouse, and then everything else fades in (including yet another advertisement for Google Chrome, and an ugly one at that):

Yes, we’ve seen this kind of hypocrisy before…

UPDATE (1/7/2010): Google has now changed the way the homepage loads, with the Nexus One ad faded out and only appearing with the other content when the use moves the mouse. I’d be curious to see what kind of discussions led to the original design, and this revert.

[Minor update: crossed out "shilling", since it is Google's own phone]

Sometimes, however, that trust can be misplaced.

Recently, a student in one of my classes gave a presentation on Google, and proceeded to explain how Google ranks search results using an algorithm called…..PigeonRank:

Why Google’s patented PigeonRank™ works so well

PigeonRank’s success relies primarily on the superior trainability of the domestic pigeon (Columba livia) and its unique capacity to recognize objects regardless of spatial orientation. The common gray pigeon can easily distinguish among items displaying only the minutest differences, an ability that enables it to select relevant web sites from among thousands of similar pages.

By collecting flocks of pigeons in dense clusters, Google is able to process search queries at speeds superior to traditional search engines, which typically rely on birds of prey, brooding hens or slow-moving waterfowl to do their relevance rankings.

When a search query is submitted to Google, it is routed to a data coop where monitors flash result pages at blazing speeds. When a relevant result is observed by one of the pigeons in the cluster, it strikes a rubber-coated steel bar with its beak, which assigns the page a PigeonRank value of one. For each peck, the PigeonRank increases. Those pages receiving the most pecks, are returned at the top of the user’s results page with the other results displayed in pecking order.

PigeonRank, of course, is a hoax, part of Google’s 2002 April Fool’s Day joke. But how did my student fall for it in 2009?

Simple. He trusted Google.

The first result when you search Google for “How does Google work?” is a link and a blurb purported to describe precisely that:

My student clicked on the link, read, and digested the information. He trusted Google.

Sure, a bit more information literacy might have tipped him off that this was a joke, but, like many folks, he had no real clue how Google works and simply trusted the result. (At the bottom of the page is a disclaimer that it is just a hoax, but he must not have seen it.)

So, making this a teaching moment, we learn from this experience that:

1. Many people still do not understand how Google works.
2. Many people trust Google search results and believe that what appears first is the correct information.
3. We need to teach information literacy as a core competency for all students.
4. We need to think about whether Google has any kind of responsibility to ensure hoaxes (especially of its own making) are marked as such within Google results. (I’m not necessarily suggesting this, but it makes for a good classroom discussion)

UPDATE: The minds at Crooked Timber have a nice discussion of this episode…

UPDATE: This item has been picked up over at MetaFilter.

Aesthetically, this isn’t without its charm, and we all know how anal Google can be regarding the minimalist design of its homepage.

But it took so much effort to get Google to finally add a mere 7-letter hyperlink to its privacy policy on the homepage, it pains me to see it disappear again….

UPDATE: It has been suggested to me that having the privacy policy link “fade-in” might actually draw more attention to it than it simply being visible from the onset. That’s a testable hypothesis that I hope one of my quantitative-leaning colleagues might tackle.

Yet, the fact remains that in Google.com’s natural state, the link is invisible. A user can visit the site, simply start typing search terms, and click return to view the results. Thereby, preserving Google’s minimalist aesthetics, and keeping the links tucked out of view. That might be cool web design, but it is poor privacy compliance.

Speaking of, I wonder what the State of California will think about this, since it was California’s law requiring a privacy policy be “conspicuously” placed on a website’s homepage that finally compelled Google to add the link in the first place. If a user has to interact with a page in order to make the link visible, is that sufficiently conspicuous?

Last week, however, this all changed. Google announced two new “features” in Latitude: Location History and Location Alerts.

Location History allows users to opt-in to having Google keep a history of their locational data tracked by Latitude. Only you can see it, and you can remove items from your history, which is great. But for everyone who activates this service, there’s now a log in Mountain View of everywhere your cellphone has been, a log that could be shared with third parties in according with its privacy policy.

More people might activate Location History when they learn about Location Alerts, a service that notifies you if a friend happens to be nearby. The beauty of Location Alerts is that you won’t be altered when people are simply engaging in their routine activities (ie, you won’t be alerted every time your coworker sits down at their cubicle across from you) . Instead, it “learns” what users’ “normal” locations are, and only notifies friends if they are nearby in an unusual place or time. To make this work, you need to have Location History activated, and in the process, Google is able to create a type of “locational profile” for each user. It is unclear whether this profile might be used for other purposes (ie, targeted advertising).

Google, of course, realizes the privacy implications of all this, and again takes some steps to help mitigate these concerns. there are FAQs for each product detailing how they work and the privacy concerns; the services are op-in; users are reminded periodically when they have Location History activated (Google should do this for all products, btw).

But all this makes me wonder: did Google plan to provide these services from the start, just with a delay? Did Google learn the lessons of Facebook, who repeatedly bites off more than it can chew as it relates to users’ privacy, and decided to launch Latitude without these features, thereby winning the praises of privacy advocates (guilty), and then strategically add them 9 months later, claiming it is simply in response to user demand?

If my fears are true, it’s not quite what I had in mind when calling on Google to engage in value-conscious design in order to protect user privacy.