In an age of leaks, forgeries, and Internet hoaxes, archivists must guard our information while keeping hackers at bay
By Elena S. Danielson
On November 1, 2010, Julian Assange, the elusive head of the organization WikiLeaks, stormed into the London offices of The Guardian to confront the newspaper’s editor, Alan Rusbridger. For months, WikiLeaks had been exclusively supplying The Guardian, The New York Times, and Germany’s Der Spiegel with hundreds of thousands of leaked classified dispatches from the fronts of the Afghanistan and Iraq wars. Editors from the three publications had been allowed to sift through the documents, assess their authenticity and veracity, then publish stories based on the material (though not before WikiLeaks dumped the documents on its website). The arrangement represented a sensational news coup, the ultimate exercise of the power of a free press. But over the course of several months, Assange’s relationship with the editors became strained. He was especially furious with the Times (in part because the paper ran an unflattering front-page story on Assange), so much so that when he provided The Guardian in October with another massive collection of documents—250,000 diplomatic cables between the U.S. State Department and its embassies—Assange demanded that the information not be shared with the Times. He had other grievances as well. And when he arrived at Rusbridger’s office in November, he had worked himself into a rage.
During the contentious eight-hour meeting that ensued, Rusbridger—along with Georg Mascolo, the editor of Der Spiegel, and Times editor Bill Keller, who had joined the conversation by telephone—expressed the journalists’ own frustrations, namely that WikiLeaks activists were sharing data files with a widening circle of outlets, resulting in a series of rogue leaks that violated the terms of the agreement. Each side accused the other of illegally appropriating what was perceived to be exclusive property. Tempers flared, threats were made, wine was poured to calm everyone down. In the end, however, the journalists and Assange recognized the irony of accusing each other of theft. As Mascolo pointed out, all of the documents had originally been obtained illegally.
The group was confronting a question that archivists face every day: Who controls information? Archivists, trained to open some government records and embargo others, are simultaneously charged with providing access to information and protecting confidentiality. From an archivist’s perspective, the larger question that arose during the WikiLeaks episode, quite apart from who got to publish what, was whether anybody was ethically justified in publishing the document cache at all. Perhaps not surprisingly, responses among archivists spanned the spectrum of public opinion. Many said that information just wants to be free, and that all efforts at censorship are doomed by the ease with which secrets can be posted these days. Others feared the disruptive consequences, intended or otherwise, of revealing confidential communications and even welcomed a counterattack by governments to take back control of information and protect it with firewalls and enhanced security. Indeed, over time, my own opinion has oscillated from the libertarian to the protectionist, ultimately coming to rest somewhere in the middle.
Reality, as is often the case, lies somewhere in this messy middle ground. What is beyond argument, however, is that when the sort of private, candid communications that were at the heart of the WikiLeaks documents are exposed to the cold, critical observation of outsiders, the context is often lost. Inside quips—comparing Vladimir Putin and Dmitry Medvedev to Batman and Robin, to name one instance—can be hurtful when laid open to the world. Gallows humor that helps soldiers get through the mayhem of war can sound cruel in the morning newspapers.
In my 27 years as an archivist, I have been appalled, for example, at the callous comments I’ve encountered in the memos of foreign aid workers, and I have had to remind myself that the huge personal risks they were taking said more than their sometimes thoughtless words. No government wants to reveal such casual comments to the world until they have aged into harmlessness, and no government wants its internal debates exposed to the public during wartime, potentially weakening confidence and swaying opinion. What price is there to pay if covert operations are revealed, even if the higher cause of democracy demands openness? As Stanford University’s facility security officer from 1996 to 2005, a position that granted me a top-secret security clearance, I read numerous classified documents that were best left to mellow and age before their eventual release.
Leaks are nothing new; only the technology has changed. Nineteenth-century peace activists greeted the advent of telegrams with euphoria, welcoming a new age of information without borders and the possibility of greater understanding between countries. The telegraph expanded the potential of the free press, and newspapers flourished as a result. It wasn’t long, however, before telegrams were selectively leaked to trigger war. In 1870 Otto von Bismarck released a rather mundane diplomatic cable, subtly edited to insult the French. The so-called Ems Telegram did not include any new information, but it provoked the map-altering Franco-Prussian War on Bismarck’s timetable. In 1917 the British intercepted and deciphered the Zimmermann Telegram, an encrypted message sent by German Foreign Minister Arthur Zimmermann to the Mexicans proposing a military alliance against the United States. The British carefully timed the telegram’s release to anger the American public and pave the way for American entry into World War I.
Stolen text on microfilm led to the unearthing of state secrets during World War II, played a role in nuclear proliferation during the Cold War, and figured in some iconic, if inadvertently comic, espionage tales. A hollow coin collected by a Brooklyn Eagle delivery boy in 1953 turned out to contain a coded message that led to the arrest of the Soviet spy Rudolf Abel. And it was a hollowed-out pumpkin where Time editor Whittaker Chambers hid undeveloped film: photographs of classified State Department documents used to convict Alger Hiss of perjury.
Radio was also instrumental in the Cold War, as leaks smuggled out of Communist countries were broadcast back to those same countries on shortwave sets. The Vietnam era saw the Xerox machine come to the fore, but no technological tool can compare to the Internet, which has allowed leaked information to be publicized on a scale both vast and unprecedented.
Was WikiLeaks the spark that lit the revolution in Tunisia that spread to Egypt, Libya, Syria, and beyond? For some time to come, historians will weigh the evidence, but the leaks were definitely a factor, helping unleash popular uprisings that have threatened and toppled regimes. Without a doubt, the Internet has made the work of traditional record-keeping much more hazardous than ever before. Websites pop up and close down, security gets tightened, hackers become ever more nimble and sly. As a consequence, the Internet will remain, like a fault line, unpredictable when it comes to the world of archival information. Those who ply their trade in its vicinity will have to tread cautiously.
Many democratic countries adhere to the 30-year rule when determining when to declassify government records and transfer them to archives—30 years being an accepted period for the passions of a particular historical moment to have dissipated. Only then can historians and journalists readily eavesdrop on the minutiae of internal governmental transactions, essentially studying the past by reading other people’s mail.
In the United States, excessive secrecy has more often than not been the norm. “Top Secret” stamps, for example, restricted scholarly access to volumes of mimeographed World War II directives for far longer than 30 years. Even after the war, the U. S. government continued to stamp routine memos and cables “Confidential” and “Secret,” providing no end date to the restrictions. Over several decades, these documents piled up in office storage closets, in the basements of retired government officials, and in public and private repositories—mountains of paper still technically off-limits.
Just before the Internet really took off, in 1995, President Clinton issued Executive Order 12958, an effort to declassify most archives after 25 years and provide a reasonable shelf life for new classified material. The order became bogged down in extensions, exemptions, and exceptions. Nevertheless, by 2006 an estimated one billion pages had been declassified—just in time for a new deluge of government emails and secret digital records. We emerged from one swamp and landed in another.
What happens when the desire for transparency collides with a legitimate need for secrecy? The American public seems to forgive a leak when it concerns public safety or tax-funded covert actions. Thus the 1971 publication of the Pentagon Papers, the history of America’s involvement in the Vietnam War—the most memorable leak prior to Julian Assange—was eventually accepted in the name of public interest. (The papers were not declassified in full until this past June, four decades later.)
Documents that protect individual privacy are an entirely different matter. The biographer Diane Wood Middlebrook used Anne Sexton’s psychiatric records, normally off-limits, to write a splendid biography of the poet. But do great scholarship and vivid writing justify a violation of medical ethics and confidentiality? The founders of WikiLeaks concede that such things as a person’s medical records should be kept secret. Even free-information zealots and techno-libertarians jealously guard their own privacy with online aliases; in hacking open the secrets of others, they employ encryption and secure communications to protect their own identities. When it comes to freedom of information, there is plenty of hypocrisy to go around, but some level of secrecy is necessary for governments, businesses, and whistleblowers.
This is where the abstract ideal of transparency in a well-functioning open society runs up against a complicated reality.
In my experience, attitudes toward transparency depend on just whose ox is being gored. I have worked with academic researchers who honestly and earnestly believe that all records should be open for examination. (Of course, they don’t want their students to have access to the deliberations of their own tenure committees.) The people who donate papers to an institution often have conflicting notions of confidentiality, as well. One donor, the son of a prominent diplomat, insisted that his father’s collection be accessible to all researchers, but when a scholar published an unfavorable article based on the archive, the incensed son demanded in vain that the papers be made available only to those who would agree to write favorably about his father. Another donor wanted to open military personnel records of foreign soldiers but was adamant that his own file be privileged.
Several prominent government officials have said that the diplomatic cables released by WikiLeaks contain nothing new, yet they maintain that the cables still put people at risk (actual fallout from the releases is difficult to assess, but several U. S. diplomats were withdrawn, and the American ambassador to Mexico resigned after comments he made in one cable infuriated Mexican President Felipe Calderón). Meanwhile, some info-libertarians contend that no one was harmed by the release of the cables yet also insist that the information unleashed violent revolutions. Contradictions inform the arguments on both sides, and they inform the way the federal government treats secrecy, too. The guardians of government secrets conveniently use certain leaks for their own purposes—the outing of CIA operative Valerie Plame comes to mind—and suppress others, for example, when Condoleezza Rice and George Tenet persuaded New York Times journalist James Risen not to write about attempts to destabilize Iran’s nuclear program. All the while, the government purports to want an airtight security system. But then, why grant hundreds of thousands of employees access to classified information, as is often the case? What kind of secret is it if half a million people are in the know?
We archivists are in a constant battle against tampering and misrepresentation. Forgery, after all, has been a flourishing literary form for at least 2,500 years. Think of the bogus Hitler diaries, sold to a German magazine for $4.8 million in 1983, or the altered documents that tripped up Dan Rather’s reporting on President George W. Bush’s National Guard service, to cite two well-known examples. Given the widespread prevalence of web-based forgeries, a “smoking gun” document from an anonymous source, though useful for a journalist, makes archivists even more suspicious these days. Even if the facts in the document are “true,” the document itself may have been falsified. And truth in a forged document is unusable. Just how easy is it to be misled by the historical record? One former cabinet-level official told me that the best way to retaliate against an enemy is to plant a damaging memo of a conversation in the filing cabinet and let it be found through a Freedom of Information Act request.
The process of examining a document’s authenticity—carefully studying and preserving clues regarding its provenance, chain of custody, and context—is called “diplomatics” and has its roots in the 17th century, when scholars sought to identify the huge number of forgeries that had been salted away in archives since the early Middle Ages. Digital documents are even easier to forge than paper ones, and untold numbers of bogus texts circulate on the Internet. Some are obvious fakes, like the digital versions of the infamous “Protocols of the Elders of Zion,” an anti-Semitic screed that has been continually reissued for more than a century. Harder to catch are authentic documents that have been tampered with. In the late 1990s, a researcher named Thomas Lowry claimed to have discovered what was very likely Abraham Lincoln’s final official directive before being assassinated: the pardon of Patrick Murphy, a Union soldier who had been court-martialed for desertion. Only recently, however, did staff members at the National Archives discover that Lowry had altered a single digit on the pardon, changing the date from April 14, 1864, to April 14, 1865, to elevate the document’s historical importance. Even the insertion of a single word, like “not,” can flip the meaning of an otherwise authentic-looking cable. Easily done, hard to catch.
Those of us who practice diplomatics in the Internet era must be versed in the field of digital forensics. What office produced the digital document in question? When did it do so? In what format, and in conjunction with what other documentation? And what sequence of equipment was it transferred on? Before digital texts can be secured with a check-sum or digital signature of some kind, they first need to be captured with all the source data (or metadata) intact, but it is precisely this information that journalists or administrators of leak sites often seek to obscure.
The editors who analyzed WikiLeaks’s diplomatic cables employed an army of experts not only to verify the reliability of the most newsworthy items, but also to redact personal names and other details that could have jeopardized lives or the success of intelligence-gathering missions. This elaborate process required people, expertise, time, and money. But redaction of any kind damages the metadata and, therefore, the authenticity of digital archives. Similarly, the anonymity provided by whistleblower websites, accomplished by rerouting and laundering messages, strips away essential provenance information. Once the damage is done, recovering an authentic version of a single altered text can be next to impossible.
Since at least 1766, Swedes have had free access to public records. This tradition of transparency is so strong that for generations a Swedish citizen’s salary has been considered public information, made available in tax calendars in libraries and now on a user-friendly website. Want to know how much money your boss makes? No problem. At the same time, Sweden has some of the strictest privacy laws in Europe (where, in general, countries are more careful with personal data than the United States is). Sweden passed the world’s first national data protection laws in 1973, a year before the U. S. Privacy Act, and Swedish laws protecting a journalist’s sources are some of the most stringent in the world. It is no accident that WikiLeaks tried to establish a home base in Sweden. Of course, a system that combines transparency and data protection demands something more than a high level of technical expertise from those who maintain it; the system needs to be part of an ethical information culture.
That sort of culture has shown signs of taking root here, too. During the tobacco wars two decades ago, when lawsuits were filed all over the country in an attempt to recover the medical costs associated with nicotine addiction, pirated copies of internal tobacco company memos arrived at the archives at the University of California, San Francisco (UCSF) School of Medicine. Tobacco company executives, under oath, had denied that their products were addictive and sought to shift the burden of medical liability to their customers. But the purloined documents the archivists received—what would come to be known as the Cigarette Papers—suggested that the executives knew more than they were letting on.
It was 1994. Newly devised graphical Web interfaces made the Internet widely accessible. Scanning technology was just becoming commercially available. In a bold move motivated by both idealism and pragmatism, the archivists scanned thousands of the pirated documents, and at midnight, July 1, 1995, they posted the documents—indexed and searchable—online, for free. The website received thousands of hits within minutes. The archivists, fastidious keepers of secrets, had determined that the public health issue at hand—presenting all the facts at precisely the time when public policy was being formulated—was more important than the false civility of shielding confidential, proprietary information. They braced themselves for a heavily funded legal assault.
“The legal challenge was very, very time consuming,” recalled Karen Butter, the director of the UCSF library and archives, “both in working with our legal counsel and in giving depositions. But I would not have continued the fight if I didn’t feel strongly about freedom of information.”
The similarities between the UCSF archivists and WikiLeaks begin and end with both groups having posted purloined papers on a website. The differences, however, are numerous. Preserving scientific medical information is the mission of the UCSF archives. The staff had a great deal of experience with scientific records and knew how to collect, authenticate, catalog, and reference such data. The metadata of the pirated materials were consistent and well documented. Any complaints about privacy violations were carefully reviewed. The papers, presented without publicity or commentary, contained vital health statistics—they did not affect national security. The archivists were transparent and open in their own work. No hypocrisy there.
Establishing trusted archives that balance freedom of information and privileged data is not easy, but it is possible. And, I would argue, it is possible to construct reasonable protections for whistleblowers, for those cases in which leaks are released in the public interest. A reliable information infrastructure—an ethical mandate in a democratic society—depends a lot on carefully crafted technology. To be sure, maintaining an ever-growing mass of digital information and migrating it to other platforms as technologies change is a daunting task: difficult, constantly evolving, and very costly. Still, traditional records managers appreciate the need to upgrade security systems as hackers increasingly threaten proprietary data. Overcoming this challenge, in all its blooming complexity, means staying a step ahead of the hackers and forgers of our day, who grow cleverer and more elusive all the time. Of course, balancing transparency and confidentiality has been the archivist’s challenge since the day writing was invented. Julian Assange simply escalated those stakes. In that sense, WikiLeaks just may have been good for business.
Elena S. Danielson is the author of The Ethical Archivist and worked in the Hoover Institution Archives at Stanford University from 1978 until her retirement in 2005.
Comments are closed for this post.