Nearly two dozen people—a group containing two women, two children, and several armed men wearing military fatigues—come into view. According to our narrator, the leader of the group is Master Corporal TchoTcho. It gets ‘bloody’ when he is around, we are told.

TchoTcho shouts at one of the women. ‘You’re going to die, Boko Haram!’ He drags her by the hair as she clings to a young girl. The women and children are led to a clearing where they are forced to kneel and are then shot twenty-two times in the back.

This horrific video surfaced on Facebook in July 2018 and quickly went viral, spreading across social media platforms. The person who filmed it was unknown, as was how it was released, and by whom. At first viewing, it seemed impossible to determine the location. Apart from the individuals, all the video showed was a dusty path surrounded by low buildings, shrubs, and trees. Comments on social media suggested Cameroon as the locus, though the government of Cameroon’s Minister of Communication quickly dismissed the video as ‘fake news’ and a ‘gross misinformation whereby the facts have nothing to do with the work of [Cameroon’s] defense and security forces’.¹

Human rights advocacy organizations—ranging from large non-governmental organizations (NGOs) such as Amnesty International and Human Rights Watch to smaller groups like the Syrian Archive—have become increasingly familiar with videos such as this that appear to depict human rights violations. They now devote considerable resources to investigating such incidents and other evidence that appears on line and off. In the video clip described above, for example, the horror of the incident caught Amnesty International’s attention and it dedicated a team to studying the video (including one of this book’s editors and two chapter authors) and tracking down other leads to verify the video and determine whether Amnesty had enough evidence to argue that the video showed an extrajudicial execution carried out by members of the Cameroon military.

(p. 4) Human rights researchers watched and rewatched the video, hunting for any clues that could help solve the mystery of where and when the executions took place, and who is depicted. The men’s uniforms were compared to official photographs of the Cameroon military. The topography was analysed, the flora and fauna shown in the video cross-referenced with records of the northern part of the country. Cameroonian contacts were asked to listen to the speakers’ accents. The weapons were identified. And here was the key detail: some of the weapons—which were simply labelled as AK-47s by the Cameroon government—were unusual. One of the group was carrying what turned out to be a Zastrava M-21, a weapon that most of Amnesty International’s research team had not even heard of. This simple detail was crucial. Yes, the uniforms were from the Cameroon military (they could have been stolen, was the government’s reply). Yes, the landscape was consistent with north-western Cameroon (but also with an adjacent area of Nigeria). Yes, the men were speaking French with a Cameroonian accent (but this did not mean they were military). But the Serbia-made Zastrava M-21 is rare, and Cameroon is one of the very few countries to which this weapon is exported.² It was only by conducting this in-depth analysis over several days that Amnesty International felt confident to issue a press release identifying the perpetrators as members of the Cameroon military. With a huge reputation at stake, and human rights abusers always looking for a way to discredit human rights defenders, this type of caution is critical. But here, after careful, systematic research was conducted, was a case that could not be ignored. After Amnesty issued its press release,³ the video was picked up and reported on by major news organizations such as Reuters, the Associated Press, the BBC, and Al Jazeera. Meanwhile, the Cameroon government continued its strategy of denial.

The government of Cameroon could no longer avoid addressing the incident, and soon announced the arrest of members of the military believed to be linked to the events. Other organizations and volunteer collectives became involved. Collaboration among researchers yielded the exact location of the execution: outside a small village called Zelevet in north-western Cameroon, just a few kilometres from the country’s border with Nigeria. This discovery led researchers to travel to the region to interview possible witnesses. Several corroborated the incident. The date of the filming was narrowed to 2015. Building on Amnesty’s work and the collaborative research across the open source community, the BBC produced an in-depth exposition of the video bringing all these different elements together which itself went viral.⁴ Without such painstaking analysis and rigorous verification, the type of which is outlined in this book, it is unlikely that this video would ever have garnered an international spotlight, or resulted in bringing perpetrators to account in any form.

This single video serves to illustrate what compelled the writing of this book. Finding and using open source information available online and piecing together corroborating information to challenge official narratives are crucial to human rights work today. With more (p. 5) photographs and videos taken every day, with high-speed internet connections crossing the globe, and with social media networks increasingly available at low cost, people are sharing their experiences online at a rate never before seen. When these experiences relate to human rights abuses or violations, they can constitute information crucial to both human rights documentation and legal accountability.

Who shot the video in Cameroon may never be known. The videographer may even have been a perpetrator. This is not unusual—perpetrators regularly capture their atrocities as trophies to show colleagues, or as a form of political protest. What is clear is that, irrespective of who films a human rights violation, if the content is to be used for documenting a crime and holding those responsible to account, it must be verified. As we saw with the Cameroonian Minister of Communication in our example above, officials are quick to dismiss authentic documentation of human rights abuses as ‘fake news’. It is therefore essential that those monitoring human rights violations and abuses around the world are able to convincingly verify the content of the information in question. They must be convinced that they are on solid ground, and that a piece of information depicts what it claims to depict, including the when and where. This is especially true if monitors hope to secure justice through courts. Providing guidance on and insight into how this can be achieved is our objective in creating this book.

Digital Witness brings together leading experts on open source research, in order to share the methodologies used to discover and verify content, to highlight key factors to consider when undertaking this type of research, and to discuss how open source methods can contribute to documenting human rights abuses and bring perpetrators to justice. These methods are relevant to everyone with an interest in discovering the truth—from journalists, to lawyers, to human rights activists, and concerned citizens.

Open source information—publicly available information that anyone can obtain by request, purchase, or observation—has been a valuable resource for a long time. But it is the volume of content available and the speed of its transmission and relay that has radically changed human rights organizations’ ability to use open source content for advocacy and accountability, ushering in a new era of human rights investigation.

1.  The Rise of Open Source Information

The embrace of digital open source information by journalists, human rights activists, and lawyers has occurred in a largely ad hoc manner, as individuals and organizations became aware of the investigative possibilities inherent in modern communications, and began—often tentatively—to adapt their work practices accordingly.

The recent history of open source information is marked by a number of milestones. The Indian Ocean tsunami of 26 December 2004 was one of the first times that news organizations received large volumes of what we now call user-generated content—video and photographs captured by the public at large and shared on social media. At the time, ‘smart’ telephones with built-in multi-megapixel cameras did not exist, but hand-held video cameras were common. It was these cameras that tourists in Thailand, Sri Lanka, and Indonesia used to capture the tsunami’s immediate impact and subsequent devastation as the giant waves hit the shores. Soon after, events of the 2007 Saffron Revolution in Myanmar and post-election protests and violence in Iran in 2009 were captured on grainy, low resolution (p. 6) cameras showing how governments were repressing protest on their streets—violently in many cases. In 2009, a group of journalists based in Ireland created Storyful, the world’s first social media news agency, which developed new methods for effectively and efficiently monitoring social media to detect outbreaks of violence around the world and to get that original content into the hands of media. Then came the Arab Spring, which swept across several countries in northern Africa and the Middle East, starting in Tunisia in 2010 and spreading to Libya, Egypt, and the Syrian conflict. A landmark in the mainstream media’s recognition of the utility of open source information came in 2015 when The New York Times used nine images sourced from Instagram for a front-page story on snow blizzards in New York City.⁵ However, while the use of open source information was increasing, few knew how to verify such content. The field of practice was, at this time, a ‘Wild West’,⁶ with verification often an uncomfortable afterthought.

In 2014, the European Journalism Centre published ‘The Verification Handbook’.⁷ The handbook was the first to set down a methodology to tackle a question that had been plaguing journalism since 2009: in a world in which the use of Facebook, YouTube, Twitter, and other platforms was exploding, how could news organizations leverage the power of content shared on these networks without making mistakes? The Verification Handbook was a first attempt to address this challenge, and it remains one of the ‘go to’ guides for verifying social media content.

Beyond the journalists who pioneered open source methods, two other communities increasingly began to draw on open source information: human rights activists and human rights lawyers. Like journalists, lawyers and activists are fundamentally concerned with telling people’s stories⁸ and rely on information shared by others to do so. It was almost inevitable that human rights lawyers and researchers would also turn their attention to how they could harness the power of social media, global internet connectivity, and the cheap image sensors being built into mobile telephones and other cameras to collect evidence for human rights advocacy and accountability.

A new challenge was also emerging in the field of international criminal law. In 2011, the International Criminal Court was about to mark its tenth anniversary, but with very few convictions to celebrate. Many of the Court’s cases had fallen apart at relatively early stages of prosecution. As part of a study conducted by the Human Rights Center at the University of California, Berkeley, researchers reviewed hundreds of pages of court records and conducted interviews to find out what was going on. One of the major problems faced by the prosecution, it turned out, was a lack of corroborating information. Indeed, judges chastised the Office of the Prosecutor at the Court for over-reliance on NGO reports—claiming that such reports did not constitute evidence—and on witness testimony that had little or no supporting documentation. Following this study, the Human Rights Center at Berkeley started working with the Court to explore how to make better use of new and emerging (p. 7) technologies in international investigations and prosecutions. In 2012, the Center hosted a workshop that brought together many of the organizations and individuals who had been pioneering the adoption and adaptation of digital technologies in amassing evidence for human rights legal cases. Human rights actors and organizations from around the world came together to discuss how user-generated content could be harnessed to strengthen human rights investigations for legal purposes, with a particular emphasis on gathering corroborating information from smartphones and social media. At that meeting were large human rights organizations, such as Human Rights Watch, Amnesty International, and Physicians for Human Rights; former investigators and prosecutors from the criminal tribunals established for Cambodia, Rwanda, and the former Yugoslavia; groups that worked with big data, such as researchers at Benetech (now at the Human Rights Data Analysis Group); forensic experts from the International Commission on Missing Persons and the Netherlands Forensic Institute; experts in remote sensing; and groups that relied on other forms of scientific information. Several NGOs present were at the forefront of efforts to train citizens around the world to collect information for legal cases. The NGO WITNESS, for example, launched its Video as Evidence programme soon after with the goal of strengthening the quality of citizen video for court purposes, and the Women’s Institute for Gender Justice was arming women with cameras to document gendered crimes.

The meeting was followed by a series of others, hosted variously by the Human Rights Center at UC, Berkeley, the Center for Human Rights Science at Carnegie Mellon, and other academic institutions. In 2015, the Human Rights Centre at the University of Essex launched its Human Rights, Big Data, and Technology Project to consider the challenges and opportunities presented by big data and associated technology from a human rights perspective—including the opportunities and challenges of using open source information.

These efforts are intended to ensure that technology can be developed and used to advance human rights. Not only are projects such as these resulting in fuller evidentiary records for courts, but they are also contributing to an international protocol on open source investigations that aims to improve the use of open source investigations for legal practice. For example, in 2016 the International Criminal Court began to turn such academic discussions into action by using open source methods to support the investigation of the Al-Mahdi case, which concerned the destruction of cultural heritage property in Timbuktu, Mali. Another milestone was reached when, in 2017, the Court issued its first arrest warrant based primarily on evidence collected from social media posts. The warrant was issued for Mahmoud Mustafa Busayf al-Werfalli, a Libyan commander whose followers had posted videos appearing to show him carrying out or ordering extra-judicial executions in Benghazi. And then, in 2018, the United Nations Independent International Fact-Finding Mission on Myanmar cited Facebook posts and videos in its calls for Myanmar’s military leaders to be investigated for genocide, and other crimes.

The human rights community’s embrace of this new, wide range of open source information comes at a time when physical access to sites of interest is becoming increasingly difficult, whether due to security concerns or diplomatic constraints. To give just a few examples: the Syrian conflict has taken the lives of countless Syrian and international reporters and human rights workers since it started in 2011; two UN human rights investigators were murdered in the Kasai-Central province of the Democratic Republic of Congo (p. 8) in 2017; North Korea has refused to cooperate with any form of human rights investigation within its territory; and human rights defenders in the Philippines were labelled as ‘terrorists’ by the government of President Rodrigo Duterte in 2018. Governments frequently deny visas to outside journalists and human rights workers—or put those working locally in pre-trial detention for indeterminate durations. In light of its potential to circumvent the risk to life or the loss of physical access, the use of open source information has become increasingly enticing.

While journalists, lawyers, and human rights advocates often experience similar challenges in using open source information, those challenges—while overlapping—are not the same. Typically, journalists are interested in stories that will interest their audience; human rights researchers are interested in those that depict human rights abuses; and lawyers in those that violate the law. These do, of course, intersect—but not always. While a journalist needs to get a story out as quickly as possible, the human rights researcher’s task is often slower, and the lawyer’s slower still. Lawyers also need to verify content to the highest of standards, documenting chains of custody and preserving content that may need to be held for decades before being used in court. Human rights advocates operate somewhere in a middle ground, occasionally working with courts in mind, and at other times trying to bring urgent and immediate attention to a story in order to help combat ongoing abuse.

Addressing the broad need to discover, verify, and archive—and the disparate concerns of these three, overlapping communities—is why we brought this book into being. Most of the resources that are designed to teach people how to find and use open source content are aimed at journalists. Yet verification has become a critical skill for many other human rights-oriented communities. There has been no comprehensive source available to introduce students and practitioners to the broad array of skills that are required in this new world of investigations. While many universities teach how to conduct field-based human rights investigations, there is a gap in university coursework that this book—aimed at academic audiences, students, and practitioners—hopes to close. We want to arm the next generation of lawyers, journalists, sociologists, data scientists, activists, and researchers with the cutting-edge skills and insights needed to work in an increasingly digitized and information-saturated environment. With this work poised to explode in importance and prevalence over the next few years, human rights organizations need to ensure that their staff—and academic institutions need to ensure that their students—are equipped to tackle these modern-day challenges.

2.  Definitions

3.  A Guide to This Volume

This edited collection is written by distinguished practitioners and academics who are pioneers in the use of new technologies in human rights research and investigation. While each chapter is intended to stand alone, we have organized the text in several sections: First, the book situates open source investigations in an historical, social, and theoretical context. Next, it covers the logistics of discovery, verification, and archiving. It then discusses the possibilities and limitations of using open source information in human rights monitoring and documentation, and suggests how future developments in open source information technology may affect human rights work.

In Chapter 1, Christoph Koettl, Daragh Murray, and Sam Dubberley discuss the history of using open source information in human rights reporting, and then, in Chapter 2, (p. 11) Alexa Koenig discusses the history of open source investigations for legal practice. Lindsay Freeman next analyses how to use open source digital content in prosecuting grave international crimes, discussing some of the lessons the Office of the Prosecutor at the International Criminal Court has learned in grappling with the new, digital information environment. In Chapter 4, Ella McPherson, Matthew Mahmoudi, and Isabelle Guenette Thornton discuss some of the big-picture social considerations that underlie this field of practice: whose stories does open source information privilege and whose does it obscure? Who are the information workers who have access to this kind of content, and whose labour is minimized or excluded? Scott Edwards then outlines some of the current and future challenges that underlie how to use open source investigations in human rights practice.

Part II focuses on using open source information in practice. Paul Myers provides an overview of methods for conducting discovery using open source techniques, including how to glean material from social media sites like Facebook and Twitter. How might a researcher construct an investigatory plan that would help ensure the systematic collection of relevant information? In Chapter 9, Aric Toler discusses how to verify that content, trust being the currency of all human rights researchers. How, for example, can we establish that a video or photograph shows what its originator or sharer claims to portray? Micah Farfour tackles the use and analysis of satellite imagery and other remote sensing data for verification and documentation. Yvonne Ng outlines how to archive open source information appropriately, while Jeff Deutsch and Niko Para build on her work to explain how information can be scraped from the internet and archived en masse.

In Part III, the book shifts from the pragmatic considerations underlying open source investigations to issues of ethics and security, whether physical, digital, or psychosocial. In Chapter 11, Zara Rahman and Gabi Ivens discuss the ethical questions that should be considered when using open source information in human rights research. Sam Dubberley, Meg Satterthwaite, Sarah Knuckey, and Adam Brown summarize their research into secondary trauma—the psychological or social stress that can emerge from experiencing the first-hand trauma experiences of another—which is a risk for all human rights practitioners but becomes especially acute when looking at large volumes of graphic footage, and provide practical suggestions for how to build resiliency. In Chapter 13, Joseph Guay and Lisa Rudnick discuss the digital and physical security concerns specific to this area of practice, pulling from their analysis of a unit that conducts open source investigations for both human rights reporting and case building.

In Part IV, we contemplate the future of open source investigation in the field of human rights. Fred Abrahams and Daragh Murray provide a forward-looking perspective on how open source information can be responsibly and effectively harnessed to strengthen advocacy and increase awareness of human rights abuses globally. And in Chapter 15, Alexa Koenig and Lindsay Freeman outline minimal standards and best practices for adapting open source methodologies when researchers or legal investigators hope to maximize the value of their work for courts.

When marching two women and two children down a dusty path in north-western Cameroon, TchoTcho and his followers probably never imagined that their crimes would be witnessed by the world. A video recorded by a simple camera built into a mobile phone and shared through the internet made this possible. The video inspired the human rights community to rise up and challenge the executioners’ actions. But human rights researchers were only effective because they were trained to interrogate the video’s veracity. Such methods must now become an essential part of every human rights researcher’s toolkit. We hope this book helps to make that potential a reality.