The Hybrid Media System

Home > Other > The Hybrid Media System > Page 21
The Hybrid Media System Page 21

by Andrew Chadwick


  The discourse that emerges from some of the professional journalists’ accounts, particularly that of Keller, is of the need for professional news organizations to tame an “anarchist” WikiLeaks hell-bent on publishing everything in its possession. Yet there is little hard evidence that WikiLeaks had a cavalier attitude to sensitive information at this stage. As the Guardian journalists themselves admit, by the time of the embassy cables Assange was as eager as the professionals to be selective in what was published and to redact the documents, not only to protect sources but more pragmatically to avoid provoking unnecessary hostility to the project among professional journalists. Indeed, during the Afghan war documents release earlier in 2010 WikiLeaks had removed a batch of fourteen thousand files because they may have contained identifying information (Domscheit-Berg, 2011: Ch. 15, para. 25). The Iraq war logs of October 2010 were also heavily redacted (Leigh & Harding, 2011: Ch. 8, para. 25) And yet the perception that WikiLeaks was an irresponsible organization and “not journalism” was very widespread. According to a content analysis of newspapers from November 14, 2010 to January 28, 2011, some 60 percent of stories misleadingly referred to the “dumping” of “250,000” cables (Benkler, 2011). In fact, by the end of December 2010, a total of just 1,942 had been released.

  In the day-to-day processes of news production the worlds of WikiLeaks and the journalists were sometimes uncomfortably far apart. The Guardian’s David Leigh and Luke Harding reveal that there was a good deal of muddling through in handling data and digital tools. The sheer size of the two hundred million-word embassy cables database forced Guardian journalists to run keyword searches using TextWrangler, a rudimentary piece of Macintosh text editing software (Leigh & Harding, 2011: Ch. 11, para. 16). Terms that were too generic would produce thousands of results, so they would sometimes resort to searching for unusual phrases in the hope of hitting some bizarre description of events. In discussing how to publish stories based on the Afghanistan documents the Guardian’s journalists spoke of coverage in the anachronistic language of paper and the old news cycle: “14 pages, on the day of launch” (Leigh & Harding, 2011: Ch. 8, para. 20). Far more readers would encounter this material on the paper’s website, where “page” space was not a constraint. As if to force this digital media logic home, the official launch of the cables stories was actually scooped by an anonymous individual named freelancer_09, who tweeted page scans of all of the major headlines from a paper copy of Der Spiegel he or she picked up from a batch left by mistake at Badisher Bahnhof station on the Swiss-German border (Leigh & Harding, 2011: Ch. 15, para. 14).

  But we also must recognize that, however much they struggled to come to grips with WikiLeaks’ infrastructure and working practices, the journalists offered their own network resources. They improvised their own form of secure transnational network infrastructure, meshing their practices with those of WikiLeaks. E-mail was out of bounds; Skype video calls were used instead. During these sessions the code numbers of relevant cables would be silently held up to the camera as a means of evading interception. Access to the cables was also provided over encrypted VPN connections. The Guardian’s cables team were given temporary “burner” pay-as-you-go cell phones as a means of evading wiretaps (Leigh & Harding, 2011: Ch. 14, paras. 19–21).

  Journalists adapted to the demands imposed by the huge volumes of data, some of which required great effort to verify. A team of experienced war reporters was assembled to undertake these tasks, including Jonathan Steele and James Meek of the Guardian, Eric Schmitt of the New York Times, and John Goetz and Marcel Rosenbach of Der Spiegel (Leigh & Harding, 2011: Ch. 8, para. 16). Their team also included Alastair Dant and Simon Rogers, whose role was to create compelling visual displays for the website, integrating leaked data with temporal and locational information.

  The news organizations also provided a further set of resources: legal expertise, legitimacy, and widespread recognition among their publics. This helped the partnership as a whole resist political pressure in the run-up to publication. Two days before the launch, Guardian editor Alan Rusbridger fielded a conference call from senior figures in the U.S. administration, including the U.S. Assistant Secretary for Public Affairs Philip J. Crowley, Secretary of State Hillary Clinton’s private secretary, and members of the Department of Defense and the National Security Council. The administration’s aim was to establish precisely what was about to be leaked. Rusbridger conceded the broad themes of the first three days, but no details. Georg Mascolo, the editor-in-chief of Der Spiegel, took a similar call from the U.S. ambassador to Germany (Leigh & Harding, 2011: Ch. 14, para. 56).

  This was in stark contrast with what happened to WikiLeaks. Assange wrote to U.S. ambassador Louis Susman in London to ask that the U.S. authorities produce specific examples of how publication of the embassy cables might put individuals in danger. The response came from Harold Koh, a legal adviser to the State Department. Koh ignored Assange’s request, declared that the cable leak was “provided in violation of U.S. law,” and demanded that WikiLeaks return the stolen files (Leigh & Harding, 2011: Ch. 14, para. 65).

  These episodes are highly revealing. In contrast with WikiLeaks, the senior journalists had much experience of bargaining with elite government sources. Their editors possessed the legal resources that would enable them to craft stories to avoid provoking lawsuits. The news organizations also had clout: all of the partners were revered national institutions in their own right. Any decision by the governments of the United States, Britain, Germany, France, or Spain to attempt to suppress publication by the newspapers would have instantly provoked outrage among a significant section of each country’s population. On the other hand, suppression would have proved futile in any case because the stories were scheduled to appear simultaneously elsewhere—on the WikiLeaks website. The likelihood of concerted action by five governments was minimal. Coordinated collective action by the newspapers, WikiLeaks, and its distributed networks was a safer bet.

  The power of the traditional news organizations was thrown into sharp relief when Assange was first accused of alleged sexual misconduct in Sweden, during the publication of the embassy cables. Journalists instantly faced a significant problem. Even though Assange was a partner in their efforts, to downplay a potential personal scandal would have threatened their credibility. The New York Times started to shift their emphasis and ran a front page extended profile of Assange dealing in some detail with the sexual allegations. The Guardian and the other partners eventually followed suit, as Assange’s personal character began to emerge as a strong part of the WikiLeaks narrative and an opportunity on the part of the Guardian and its partners to report a new human interest angle in the story.

  This shift in power relations was exhibited during a fraught and revealing episode a few weeks before the embassy cable stories went public. Despite Assange agreeing that he would grant the press partners exclusive access (Leigh & Harding, 2011: Ch. 13, para. 12), WikiLeaks had itself started to leak. The entire embassy cables database had been acquired by Heather Brooke, a London-based freelance journalist and freedom of information campaigner. This led to a confrontation at the Guardian’s headquarters on November 1, 2010. During this meeting, involving Assange and his legal team, the Guardian’s senior editors and their lawyer, Der Spiegel and Guardian journalists, and the New York Times’ Bill Keller (by phone), the interdependence among these actors was played out in vivid fashion.

  Assange was angered by the hostile profile of him in the New York Times and now wanted the paper to be kept out of the deal. He mentioned potential collaboration with other papers, including the Washington Post. Meanwhile, the Guardian had been negotiating with Heather Brooke in an attempt to bring her into the fold to eliminate the risk that she might take her copy of the cables to another paper and spoil their scoop. The Guardian responded to Assange with the argument that the cables had in fact already been leaked and neither they nor the New York Times was now dependent on WikiLeaks for the data required to run
their cables articles. But the Guardian could not be sure that they would be able to secure Brooke’s compliance, nor would they have the WikiLeaks web publishing infrastructure to guarantee the worldwide availability of their evidence. Assange therefore still had power resources, derived from digital media, to deploy. Assange then asked that the embassy cables partnership be widened to include El País and Le Monde and that the New York Times publish on their front page his response to their unfriendly profile. Though Bill Keller refused Assange’s request to have his say, the meeting did agree to include the Spanish and French papers. The original partners’ exclusive was therefore diluted, but Assange failed to exclude the New York Times from the deal because the Guardian later passed the cables over to Keller in any case. This was, therefore, a victory through compromise for all involved. What follows is the revealing “10 bullet point” note on which Guardian editor Rusbridger agreed with WikiLeaks:

  Publish on Nov 29 in a staggered form.

  Run over two weeks or more up to just before Xmas.

  Exclusive to G, NYT, DS (plus El País and? Le Monde).

  Subject matter to be co-ordinated between partners and to stay off certain issues initially. No veto to anyone over subjects covered over whole course of series (post-Jan). WL to publish cited documents at same time.

  After Xmas the exclusivity continues for one more week, starting around Jan 3/4.

  Thereafter WL will start to share stories on a regional basis among 40 serious newspapers around the world, who will be given access to “bags” of material relating to their own regions.

  G to hire HB [Heather Brooke] on an exclusive basis.

  If “critical” attack on WL they will release everything immediately.

  If material is leaked to/shared with any other news organization in breach of this understanding all bets are off.

  If agreed the team will commence work on a grid of stories for the first phase (Leigh & Harding, 2011: Ch. 11, para. 40).

  The terms of this deal crystallized the interdependence among WikiLeaks and the professional news organizations. The first embassy cables articles went ahead as planned, save of course for the Der Spiegel accidental Twitter leak, on November 28, 2010.

  FROM THE NEWSROOM TO THE NETWORK

  This marked the end of an episode of collaboration, but in December 2010 WikiLeaks’ polymorphous network resources came to the fore again in the immediate aftermath of the embassy cables launch, providing symbolic and physical reinforcement to the professional journalists. There erupted a sprawling, symbolically charged cyber conflict, pitting activists engaged in electronic civil disobedience against a group of politicians, public authorities, and private companies intent on weakening WikiLeaks’ organizational capacity.

  Faced with a series of massive distributed denial of service attacks originating from botnet armies of virus-infected computers assembled by hostile hackers, including one named “The Jester,” WikiLeaks moved its main website to Amazon’s web services cloud storage—the largest platform of its kind in the world. The cables website, meanwhile, remained on a server hosted in France. The denial of service attacks on WikiLeaks main site intensified, their sources becoming more diverse.

  As with all such episodes, it was impossible to identify the origins of these hacking attacks with certainty, but they jelled with a domestic campaign by politicians in the United States to have the WikiLeaks site taken down. Senator Joe Lieberman led the charge by calling in public for Amazon to suspend its hosting and for other companies assisting WikiLeaks to end their relationship with the organization. Lieberman’s office contacted Amazon to ask if there were “any plans to take the site down?” (Slajda, 2010). In an extraordinary move, Amazon acceded to these requests and closed down WikiLeaks’ storage space on the grounds that it violated their terms of service. The company argued that customers may only publish content to which they hold the rights, referring to a clause originally designed to deter copyright infringement. But Amazon also invoked a much vaguer clause from its terms, stipulating that hosted materials should not “cause injury to any person or entity.” They claimed that the “250,000 classified documents” that they wrongly said WikiLeaks had already published would put “innocent people in jeopardy” (Amazon Web Services, 2010).

  Amazon’s decision induced a climate of fear. The precise extent to which other companies came under direct political pressure is not known and probably never will be fully known. However, some details are publicly available. Following a public statement by Osama Bedier, a vice president of PayPal, we know that PayPal received a letter from the U.S. State Department on November 27, 2010, the day before the cables stories launched. The letter stated that WikiLeaks’ activities were “deemed illegal” in the United States. In response, PayPal moved on December 4 to suspend WikiLeaks’ account, thereby choking off donations to the organization at a time when it was most likely to attract funding (Arthur, 2010). December 3 saw EveryDNS suspend the wikileaks.org domain and its associated e-mail accounts. On December 6, Assange’s Swiss bank account was frozen by PostFinance, on the grounds that he “provided false information regarding his place of residence during the account opening process.” That same day, MasterCard blocked WikiLeaks’ access to its credit card services; Visa followed suit the day after (Arthur, 2010). The Twitter accounts of Assange, Birgitta Jónsdóttir, and other WikiLeaks leaders were the subject of a subpoena from the Department of Justice on December 14. Under U.S. law, Twitter could not resist the subpoena without being prosecuted, but it made the order public after successfully applying to have its details “unsealed” and sent to those whose accounts were under investigation.

  This unprecedented series of events led to “Operation: Payback” and “Operation Avenge Assange”: decentralized online campaigns of politically motivated retaliatory hacking attacks aimed at the websites of the public organizations and companies that were suppressing WikiLeaks. The action was minimally coordinated by a loose, leaderless, memberless, and constantly shifting transnational collective of around ten thousand hacktivists named Anonymous, some of whom were associated with the “doing it for the lulz” culture of the libertine and irreverent web forum, 4Chan (Shapira & Warrick, 2010). Operation: Payback had started in September 2010 as an online protest against the Motion Pictures Association of America’s attacks on the file-sharing site, the Pirate Bay (Coleman, 2011). To discuss tactics, Anonymous used Internet Relay Chat (IRC) channels, PiratePad (a web-based text editor that provides for simultaneous real-time document writing), and polls embedded in online Google forms. Within a few days, more than fifty thousand had downloaded the desktop software that allows a user to take part in the denial of service attacks, but Anonymous almost certainly used networks of virus-infected “zombie” computers as well (J. E. Dunn, 2010). Media reports at the time suggested that Anonymous’s support for WikiLeaks was a new development, but in fact the group had played important supporting roles in the past. In 2008, for example, they had assisted with organizing website material for a leak detailing secret aspects of the Scientology movement (Domscheit-Berg, 2011: Chapter 3, para. 22).

  This was, in part, a phony war, and reliable reports are thin on the ground. Anonymous cultivates an air of secrecy, describing itself simply as an “internet gathering,” but MasterCard’s website was briefly the victim of a successful distributed denial of service attack on December 8. It also appears that Anonymous briefly closed the website of PostFinance, Assange’s Swiss bank, and caused some disruption to the Swedish government portal (Bloxham, 2010). ABC News reported that Sarah Palin had contacted them to allege that her website, SarahPAC.com, and her personal credit card details had previously been targeted by Anonymous hackers (Tapper, 2010). A number of new search engine interfaces to the cables database were launched, such as cablegatesearch.net. In addition, “Operation Leakspin” was launched by, among others, users of the popular news sharing site Reddit: volunteers posted links to the cables and readers were encouraged to “vote up” those that
were most important (Operation Leakspin, 2010). Anonymous’s Facebook page was suspended on December 8 for violation of the company’s terms of service (Bosker, 2010). And, in yet another bizarre twist of organizational hybridity, Anonymous itself issued a series of carefully worded formal press releases outlining that the denial of service attacks on PayPal, which it admitted were taking place, were electronic civil disobedience not aimed at critical infrastructure but at the “public face” of the companies and organizations (Anon Ops, 2010).

  Hacktivism in itself is nothing new and goes back to the mid-1990s (Chadwick, 2006: 114–143), but the intensity and visibility of these reactions and counter-reactions to the cables leaks was new. This was part of an ongoing public drama in which WikiLeaks’ networks of affinity were mobilized, mostly in support of Assange, but partly in support of the collaboration between WikiLeaks and the professional journalists. These actions were an essential aspect of the hybrid media system. They were a show of strength of sorts, by the members of an online “anti-leader” network who had become politicized and who were willing to take personal risks in order to demonstrate their support for transparency, freedom of expression, and the principle of whistle-blower anonymity. And those risks were very real: in July 2011 the U.S. State Department and the British and Dutch police announced that sixteen participants in the Anonymous denial of service attacks on PayPal had been arrested and charged with crimes (U.S. Department of Justice, 2011).

 

‹ Prev