Open Networks, Closed Regimes

This page documents a single class session on worldwide Internet restrictions taught by Joshua Gruenspecht, Conor Kennedy, and Dan Ray in April 2009. We've filled it out with some background information and our reactions, in the hopes that it might be useful to someone else teaching a similar subject in the future. Special thanks go out to Caroline Nolan and Colin Maclay, who provided invaluable assistance both in the early days of planning the class and in the simulation setup. Thanks also to Andrew McLaughlin for taking time out of his schedule to join the class.

This page is a work in progress, and we encourage anyone to add to it. The wiki only permits registered users to edit it, but you can create a free, pseudonymous account using the link at the top right corner of the page. (You can copy our formatting or add your own; a template for the yellow annotations appears in an HTML comment at the top of the page source.)

Part I: Early Preparation/The Original Wiki

Problem to be solved

Our students had access to an earlier version of this wiki page. For the information in this section, unformatted text indicates material that was available at the time of our presentation, and yellow boxes like this one are annotations added after the fact think of them like the director's commentary. The actual revision of the page at the time of the class session is here

US-based internet services are used by citizens of every regime in the world. Western-based internet companies operate these services in many countries where governments routinely request or demand sensitive information about their citizens. How should high-level officers respond when their companies receive requests to turn over information which violate privacy and anonymity expectations? How can citizens, NGOs, and government actors influence the way these companies respond?

Precis

The idea of the internet service provider as a border-defying government-regulation-free jurisdiction was already dying in the 1990s, but the large-scale movement of internet services into regimes without free speech protections has raised serious concerns about managing cross-border privacy standards over the last five years. From Yahoo! turning over pro-democracy Chinese bloggers to the Chinese government, to Saudi Arabia tracking porn downloaders by pulling ISP records, to South Korea trying to arrest anonymous government critics, the problems are widespread and not restricted only to regimes that Americans are used to thinking of as "repressive." The globalization of internet services raises difficult questions: What requests for information are invasive? What kind of deference is due to local sovereignty? How can the competing demands best be balanced? How can the relevant stakeholders work toward achieving a proper balance?

The Story So Far

This section constitutes the class's common ground, in that all students were expected to read it. It is intended to situate today's issues (including our simulated crisis) within the Internet's history, and to introduce the three major proposals we expected our simulation groups to cover.

Several weeks ago, our class read John Perry Barlow’s Declaration of Independence of Cyberspace (this refers to the readings for an earlier class. It's probably not essential to assign it; this summary is plenty for our purposes.), written in 1996, in which he addresses the nations of the world and “declare[s] the global social space we are building to be naturally independent of the tyrannies you seek to impose on us.” For Barlow, perhaps the most important innovation of cyberspace was its ability to act as a great leveler: “We are creating a world that all may enter without privilege or prejudice accorded by race, economic power, military force, or station of birth. We are creating a world where anyone, anywhere may express his or her beliefs, no matter how singular, without fear of being coerced into silence or conformity.” As these lines imply, the key to this brave new world of Barlow’s was the possibility of anonymity (or pseudonymity): real-world barriers placed by normative biases or governmental controls could be hurdled by individuals joining online communities to express their provocative thoughts. Secure in the knowledge that their identities were divorced from their real-world selves, citizens of every nation could create the first truly free global marketplace of ideas.

From well before Barlow’s Declaration, this idea of the Internet as an enabler of the anonymous expression of ideas was a well-established meme in the popular conception of the Internet, from Peter Steiner’s famous New Yorker cartoon to the aphorism that “The net treats censorship as a defect and routes around it.” This quote comes from John Gilmore And just as quickly, people started to raise questions about whether a completely uncontrolled discourse was desirable – Larry Lessig famously told the story (see p. 102-106) of one anonymous student who single-handedly shut down Lessig’s first online classroom bulletin board by attacking his classmates with unstinting vitriol. The techno-libertarians remained adamant that online communities would develop mechanisms to cope with such problems, and that the benefits of being able to create a discussion area open to all comers – more freedom under repressive regimes, more frank discussion for every individual – would outweigh any drawbacks.

However, the accuracy of the idea that the Internet need not respect terrestrial law was soon called into doubt. Tim Wu and Jack Goldsmith document (see Chapter 1) the 2000 case in which Yahoo! attempted to resist a takedown request from the French government. After activists discovered Nazi merchandise for sale on a Yahoo! site, they appealed to the government to enforce the laws banning such material from being sold in France. Yahoo! protested that there was no way for it to identify traffic coming from France, and that to enforce the French laws would ultimately mean enforcing the lowest common denominator of all international law in all locations. However, having lost their case in the French courts in the face of expert testimony, and facing fines of 100,000 francs per day against their new French division, the company caved and admitted that IP addresses could be used to tell from what country a visitor came. This was a blow to the notion that the Internet was inherently resistant to national laws, but it was also a blow to the idea that users of the Internet were all faceless numbers – it soon became very clear that the companies that individuals did business with on the Internet knew plenty about their users, both in the form of information submitted by those users and in the form of other incidental data.

Once those two facts were combined, it became clear that online service providers doing business internationally were vulnerable to economic threats if they failed to comply with local legal regimes by turning over nominally anonymous users. This acquired early public salience in 2004, when the Chinese government discovered that someone with a Yahoo! email account had forwarded to a US-based pro-democracy blogger an internal Communist party message warning of potential unrest which might arise from discussion of the fifteenth anniversary of the Tiananmen Square protests. China demanded that the company turn over any records relating to the disclosure of its state secrets, and Yahoo! complied without question. Chinese journalist Shi Tao was ultimately arrested and sentenced to ten years in prison. After the role played by Yahoo! was disclosed, the company became the subject of global outrage, castigated by pro-free-speech nonprofits and hauled before Congress to account for its actions. As of 2009, Shi Tao remains imprisoned. However, Yahoo! was merely the most visible of the companies involved in such disclosures – Amnesty International reports that many of the largest international online service providers were implicated at the time and continue to wrestle with these issues.

In recent years, similar cases have cropped up all over the world, from Africa to Asia to the Middle East, and even some American companies have taken part in using the disclosure-friendly laws of other countries to track down derogatory speakers. Many of these countries describe themselves as protecting themselves from the dangerous consequences of free speech: in January, South Korea arrested a blogger whose acerbic postings about government bureaucrats had supposedly undermined the population’s faith in the government’s ability to manage the global financial crisis. These governments say that free speech is always at war with social stability, as it was in Lessig’s classroom, and that they and Americans simply disagree on where the line must be drawn. At the same time, pro-democracy activists say that American companies and the American government aren’t doing enough work insisting upon the universality of American values.

The most salient legislative response has been the Global Online Freedom Act (GOFA), proposed by Representative Chris Smith (R-NJ). GOFA would require the State Department to track internet-restricting countries and then forbid US corporations from a) storing users’ personal data within the borders of those countries or b) turning over any personally identifiable information to law enforcement within those countries without first seeking the permission of the US Department of Justice. Many pro-free-speech non-profits have praised the GOFA bill as an “advance for online free expression.” Commentators have also suggested, however, that it is unworkable in practice, and would in effect require US companies to cease doing meaningful business within such countries.

Some corporations have pushed instead for a non-government-driven multi-stakeholder plan of action. Several of the largest American online service providers, including Microsoft, Google, and Yahoo!, have joined forces with the Berkman Center to develop the Global Network Initiative (GNI). The GNI lays out a set of universal principles to be referenced when making difficult human rights decisions; it then gives implementation guidance to boards considering those principles on a case-by-case basis as requests for users' data come in. Some of the Congressional critics of these companies have been mollified by the evident progress. At the same time, some legal scholars and privacy specialists (see Rotenberg quote "It's good that companies are paying more attention to the need to safeguard fundamental rights on the Internet, but this proposal lacks specific recommendations and fails to address key concerns about the surveillance and censorship of Internet users") say the GNI is not widely enough accepted or equipped with severe enough penalties to make a serious difference.

Meanwhile, countries wary of indiscriminate speech have started to look at alternatives of their own. This past year, the International Telecommunications Union (ITU), a specialized agency of the United Nations that deals with the regulation of international telecommunications issues, opened up a study group on Question 6-17 (Q6/17). The Q6/17 group was chartered at the request of a state-owned Chinese telecommunications firm to investigate the feasibility link to .doc of retrofitting the internet with Internet Protocol (IP) traceback technology. IP packets, the individual "envelopes" used to send information across the internet, are difficult, if not impossible, to follow back from their destination to their source. IP traceback technology would use new software and hardware to retrofit tracing mechanisms onto the internet as built, so as to make it easier to find the sender of a given message. The proponents of Q6/17 suggest that it could be used to make it more difficult to remotely infect computers with viruses or make anonymous threats online. Privacy specialists, however, suspect that the major purpose is to crack down on unwanted speech.

Required Readings

In the spirit of a class in which different groups were meant to represent different constituencies with different backgrounds, we decided to move away from using the readings to get the class on the same page. Aside from the text above, which was assigned as introductory material, each group was assigned individualized readings to explain "their" point of view. As a result, we removed the Required Readings section of the page entirely, archiving it, as seen below. The Additional Readings section remained in place, for those students interested in a wider exploration of the issues. The individualized readings were emailed individually to the class groups, as discussed in Part II.

Early brainstorming on this topic

Additional Readings

(Some of these and additional readings will be broken down by simulation group)

Audio:

• http://rconversation.blogs.com/rconversation/2008/07/on-the-media-in.html (Episode of On The Media w/ Rebecca MacKinnon)

Video

• http://www.youtube.com/watch?v=PMorzIBHmrA (Conversation about GNI on CSPAN w/ Michael Samway and Julien Sanchez)

Text

“Just the Facts” Reporting

• http://news.bbc.co.uk/2/hi/technology/7696356.stm

• http://www.nytimes.com/2008/10/28/technology/internet/28privacy.html?_r=1&oref=slogin

Critiques

• http://writ.news.findlaw.com/ramasastry/20081212.html (Anita Ramasastry, Associate Professor at University of Washington School of Law)

• http://www.rsf.org/article.php3?id_article=29117 (Reporters without Borders, an NGO that was part of the talks but didn’t sign onto the GNI at the end)

• http://www.eff.org/files/filenode/gni/signon_letter.txt (Electronic Frontier Foundation, an NGO that was part of the talks and did sign onto the GNI at the end)

• http://www.thenation.com/doc/20060313/mackinnon (Rebecca MacKinnon, writing in The Nation before the GNI existed, discusses four companies' culpability and diagnoses a major flaw with GOFA)

Links

http://blogs.wsj.com/chinajournal/2008/10/28/parsing-the-google-yahoo-microsoft-global-network-initiative/

http://www.nytimes.com/2008/11/30/magazine/30google-t.html?_r=1&adxnnl=1&partner=permalink&exprod=permalink&adxnnlx=1229749063-T7qNc5xv9ZLDiLcjc8AZxQ&pagewanted=print

http://blog.wired.com/27bstroke6/2008/04/republican-hous.html

http://rconversation.blogs.com/rconversation/2007/01/google_yahoo_mi.html

http://www.circleid.com/posts/print/20081028_global_network_initiative/

http://www.guardian.co.uk/technology/2008/oct/30/amnesty-global-network-initiative

http://www.economist.com/world/asia/displaystory.cfm?story_id=12783609

http://www.huffingtonpost.com/michael-a-santoro-and-wendy-goldberg/chinese-internet-censorsh_b_156212.html

http://news.cnet.com/8301-13578_3-10040152-38.html (Declan McCullagh of CNet News on Q6/17, including links to several leaked ITU documents)

Part II: Planning and Execution/The Week of the Class

Parts II and III cover the class as it played out, and represent entirely new material on this wiki. As a result, there are no yellow commentary boxes in these sections.

The lead-up to the class

The class was divided into six role-playing groups representing six constituencies with complaints about the manner in which free speech issues on the internet are currently handled. Group assignments fell naturally out of the existing divisions within the class - each role-playing group represented two of the discussion topic groups on the syllabus. Each group was emailed individually approximately seven days before the class date. The email explained the concept behind the class:

For the Anonymity and Privacy class this coming week, we're going to run things slightly differently. The class will be divided into six groups, and each group will be given an identity as a participant in the ongoing global battle over anonymous speech online, as well as a series of unique readings giving some specialized background. Then, in class, we'll introduce some general background and then have an exercise involving both y'all and our guest. See the last page of the attachment for your readings.

As the text of the email suggests, each one contained an individualized reading for the associated group. See below for the attached readings:

Group 1: The Vietnamese Government

Group 2: Vietnamese Pro-Democracy Bloggers

Group 3: The US Government

Group 4: Yahoo! (the most popular blogging service in Vietnam)

Group 5: Amnesty International

Group 6: The Chinese Government

The day of class

The room was reworked in order to remove barriers between students and facilitate discussion. The central classroom table was broken into constituent tables, each of which was placed along the outside wall of the class and topped with a name tag indicating the individual group which was expected to use that table as its base of operations. Chairs were placed inside the tables, rather than outside, to bring the groups closer to one another and the central discussion space within the classroom. (See the room plan, here. While it didn't work out exactly as pictured, the idea is the key.) Students were asked to begin the classroom seated at the tables associated with their role-playing groups so that the transition from the presentation and discussion to the simulation would be as seamless as possible.

Conor opened the class with a Power Point presentation that reviewed the basic outline of the issues at hand. This served both as a refresher of the introductory reading material and as a point of reference for the groups within the simulation.

+ Clip

After the introduction, Dan moderated a conversation between two stakeholders in the current international contretemps over networked speech control issues: Colin Maclay of the Berkman Center and Andrew McLaughlin from Google. Colin discussed difficulties involved in creating the Global Network Initiative and acquiring more signatories, especially in Europe. Andrew discussed changes that he thought would help make the GNI more useful to individual users.

+ Clip

Once everyone was caught up to speed on the tensions which would have to be navigated in the simulation, Joshua explained the simulation to the participants. Groups were asked to develop a strategy for creating a widely acceptable solution to the "Open Networks, Closed Borders" problem, to discuss their solution with other groups and acquire partners, and then to present their revised solutions to the class at large. After feedback, the solutions would be voted upon by the members of the class.

+ Clip

The first 40 or so minutes of the hour-long simulation period was scheduled as follows:

• 2 minutes: Explanation from the team
• 5 minutes: Intra-group meetings to elect a recorder/presenter and to set strategy
• 10 minutes: Inter-group meetings. Recorder/presenters stay in place while other members of the team (ambassadors) venture out to discuss the acceptability of proposals with both ambassadors and recorder/presenters from other groups.
• 2 minutes: Groups are presented with a disruptor to ensure that each one remembers that other groups have different interests: in this case, a letter offering the opinion of select members of Congress regarding the issues at hand (an actual letter; visit this site to see the text).
• 5 minutes: Intra-group meetings to report back and revise solution proposals
• 10 minutes: A second round of inter-group meetings
• 2 minutes: To write up final solutions

Each group was then given two minutes to present their final proposals to the class, and then Colin Maclay, team lead on the GNI project at the Berkman Center, offered feedback as to the viability and utility of each solution.

The proposals:

+ Clip

Colin's feedback:

+ Clip

Originally, the voting was scheduled for the last five minutes of class, but proposal presentations and feedback exceeded the available time as it was, and so voting was pushed off into post-class time.

After the Class

After class, we gave all of the participants an opportunity to vote on the proposals via email. The email, which summarized the proposals as explained in class, is excerpted here:

Each of you (individually, but remaining in your role from the simulation) should rank the solutions listed below from most to least acceptable, and draw a line between "tolerable" and "intolerable" solutions from your point of view. For example:

6 3 | 4 5 2 1 would indicate that you liked China's answer the best, could also live with Congress's, and found Vietnam's utterly unacceptable.

The carrot: Your votes will inform the final writeup that we do for the class and for the GNI team.

The stick: We will hound you mercilessly until you vote.

Thanks again, The United Nations of IIF

___________________________________________________________________

Solution 1 (Vietnam): Vietnam is willing to accept the GNI principles and best practices, but will carry them out using its own laws and and due process protections. Companies operating within Vietnam must follow all its laws and heed its national security policy.

Solution 2 (Bloggers): Vietnam should enact a DMCA-style (domestic) law that absolves bloggers from criminal liability for objectionable posts so long as they are timely removed. ISPs are also immune from liability so long as they cooperate in taking down objectionable content.

Solution 3 (US Congress): Any ICT company turning information over to a foreign government must clear that disclosure with the US government, which will apply a compelling need standard. ICT companies who turn over information without permission are subject to sanction.

Solution 4 (Yahoo!): Yahoo! proposes that individual conflicts be informally arbitrated by the individuals involved through a standardized process, and that irreconcilable conflicts be referred for binding arbitration to a special subcommittee of the WTO

Solution 5 (Amnesty): AI proposes a network of pro bono lawyers in Vietnam, who will sue on behalf of targeted bloggers when ICT companies inform AI that they have received information requests. A review board is in place to permit other nations to pressure Vietnam if necessary.

Solution 6 (China): Nations are only bound by treaties, and China is happy to opt out of anything the other groups agree on that would limit its sovereignty. Instead, it advocates the Q6/17 program, which would limit anonymity on the Internet layer.

As our team would have predicted, the teams, when asked to vote in character, voted largely for their own solutions. However, even among the second-choice solutions, no alternative was a clear favorite:

Team 1 (Vietnam): 1 6 | 2 3 4 5 1 2 | 6 4 5 3 5 2 3 4 | 1 6 5 1 | 2 4 6 3	Team 2 (Bloggers): 2 1 3 5 | 4 6 2 5 3 4 | 6 1 2 3 4 5 | 6 1 4 2 1 | 6 5 3	Team 3 (US Congress): 3 | 1 2 5 4 6 3 2 4 5 | 1 6 6 5 4 3 | 2 1 2 3 4 5 | 1 6
Team 4 (Yahoo!): 4 5 2 3 1 6 4 5 2 | 3 1 6 4 1 2 5 3 6 4 1 | 2 5 3 6 4 1 5 6 | 2 3	Team 5 (Amnesty): 5 4 1 3 2 6 5 2 4 3 1 6 4 3 5 2 1 6 4 5 3 2 | 1 6 2 4 5 | 3 1 6	Team 6 (China): 6 1 2 | 5 4 3 6 2 1 | 5 4 3 6 1 2 | 5 4 3

As the voting results suggest, China and Vietnam were most interested in each others' solutions, and the same was largely true of Yahoo! and Amnesty International. Perhaps unsurprisingly, the team representing the US Congress was hopelessly confused. There was, however, a clear loser: the team 6 (China) solution implementing Q6/17 was overwhelmingly rejected as unacceptable.

Part III: Reactions and Discussion/Suggestions to Future Class Leaders

Role of Technology

Unlike many of the other IIF classes, our team decided not to make technology tools a central part of our class plan. Feeling that the class's early experiences using the question tool and Twitter in-class had been somewhat isolating, we decided to look for an innovative pedagogical tool which would jumpstart class discussion and force students to confront each other. Ultimately, we decided that the best choice was to have the class close their laptops entirely, and that our tool would be a classroom exercise rather than a technological innovation.

This decision ultimately led to our decision to create the simulation, whose successes and failures are discussed in this section. Generally, however, we feel that the simulation was at least as successful as any network tool in involving the class - in fact, many of our voting query responses stated that class members found the exercise particularly enjoyable.