Until now. In a Jan. 28 press release, Amnesty International announced that its investigations into the Burundi killings found what appear to be mass graves that likely contain some of the missing bodies. The organization continues trying to identify other possible mass gravesites.
Clever forensic use of public satellite imagery, mobile phone videos, and more
But here’s the really big deal: how Amnesty did it. They correlated date-sequenced (before and after) commercially available satellite imagery (see the after photo above) with witnesses’ accounts and ground-level video probably taken with a smartphone to reveal areas of disturbed earth that may be the burial sites in question.
This isn’t the first time that we’ve learned via non-governmental sources about violations of human rights or international security concerns from the emerging combination of surveillance technologies, social media and the Internet. And we will be learning even more in the years to come. There has been a great deal of discussion about how these new technologies will enable governments to intrude on personal privacy and liberty.
But the reverse is also true: With these new publicly available tools, applications and much greater connectivity, governments and others will find it increasingly difficult to hide misdeeds.
Consider, for example, that public analyses of commercial satellite imagery and geo-located video have already revealed the construction of artificial islands by the People’s Republic of China in contested waters, and details of North Korean and Iranian nuclear weapons and strategic missile programs.
But that is only the beginning. Citizen analysts have produced detailed investigations of specific events occurring in crises such as the Russo-Ukraine war and the Syrian civil war. They’ve done so using the same methods, adding some very clever forensic tools. For instance, they extract details from social media (such as geo-tagged “selfies” or video posted online); employ algorithms to calculate time of day or size of objects from photos or videos using sun angles; study artillery and rocket impact craters to figure out an attack’s launch site; and track the location of military equipment or even individual vehicles with pictures, video, and commercial street level imagery.
Journalists have always been investigating governments, right?
Of course, investigating the actions of governments, corporations, and others is hardly new. Professional journalists and NGOs (especially those dealing with humanitarian relief, arms control and human rights) have long analyzed eyewitness accounts, physical evidence, and public and private documents to figure out what really happened and who was responsible for atrocities and malfeasance, despite official denials and secrecy.
But until these new technologies emerged, in virtually all cases (excepting the unauthorized release of official secrets), journalists and researchers have had to rely on governments for data such as satellite imagery, and governments are ordinarily not forthcoming, especially when it might reveal the state’s sources and methods of collection.
But governments are no longer the only ones with these data. Rapidly expanding networks of telecommunications users now enable ordinary citizens to openly collect, analyze and disseminate data that used to be closely held state secrets.
Clearly, some in the unregulated world of public source analysis are pursuing specific agendas. We should examine any data with appropriate caution and due diligence. But with that said, public source analysis is frequently more revealing and convincing than intelligence information coming from government sources. As I discussed last year, this has certainly been the case with regard to the ongoing conflict between Russia and Ukraine. The official photos and satellite imagery released by the United States and NATO have been vastly overshadowed by public source analysts such as the UK-based group Bellingcat, a transnational collective of Internet sleuths whose mission is to observe and report on what governments are doing on the ground.
How do you catch an atrocity?
Surveillance systems, especially drones, may seem to be everywhere. But major acts of violence are perpetrated around the globe daily, often with no warning, straining even the vast resources of governments to analyze them. Then there is the question of how much information governments will choose to make public.
Certainly, there has been excellent in-depth reporting using satellite imagery and other media by the State Department and other government agencies on humanitarian disasters such as the catastrophe in Darfur and the ongoing global migration crisis, some of it made public. But many people assume that states can collect and analyze intelligence data more thoroughly and more rapidly than they actually can. For instance, the U.S. State Department used the AI report on the possible mass graves in Burundi to emphasize its concerns about that country’s ongoing violence. After the slaughter was first reported, State’s messaging on Burundi included nothing substantive that relied on U.S. government sourced information. That suggests they either did not have reliable sources of data or hadn’t analyzed and released what they did have.
Trying to understand what happens in such cases becomes an exercise in forensics. Sometimes public sources have the best available evidence. Consider the tragic downing of Malaysia Airlines Flight 17 over Ukraine on July 17, 2014. Many of the countries involved directly or indirectly, including the United States, made official statements about the incident. But the Dutch Safety Board, the government body overseeing the investigation into the incident, didn’t issue its official final report until Oct. 13, 2015. Even then, the Dutch didn’t identify who shot down the plane, thereby killing 298 civilians. Dutch authorities are now conducting a separate criminal investigation, the results of which may be published later this year.
Yet by late 2014 the public source analysts — with website Bellingcat in the lead — had compiled a very convincing case about who shot down the plane and how. For a recent summary of their work, see here.
In 1999, the political scientist Ann Florini and her colleague Yahya Dehqanzada foresaw that publicly available commercial satellite imagery would “… shift power from the former holders of secrets to the newly informed. That has implications for national sovereignty, for the ability of corporations to keep proprietary information secret, and for the balance of power between government and those outside it.”
They were more prescient than they knew. That shift of power has only accelerated with the advent of globally networked social media and smart digital telecommunications devices. More and more people join the ranks of public source intelligence analysts every day. Governments like those of Burundi, indeed governments everywhere, have been put on notice: Little Brother is watching you.
Ralph S. Clem is emeritus professor of geography at Florida International University in Miami.