Investigating the Karma Massacre using OSINT
This blog post was based on research and analysis conducted as part of the final assignment for the bespoke Knowmad OSINT training for the International Masters in Security, Intelligence, and Strategic Studies (IMSISS) offered by the University of Glasgow, Dublin City University, Charles University, and Trento University.
Disclaimer: viewer discretion is advised
Increased online chatter indicates attack in Karma
At the end of April 2023, Open Source Intelligence (OSINT) analysts noted an uptick in chatter regarding a potential massacre in Burkina Faso, with reports circulating that unarmed civilians in the small northern village of Karma, Nord region, were rounded up by an armed group of assailants and systematically eliminated. In an area rife with fighting between Islamic insurgents and Burkinabe military forces, it was initially unclear which group actually perpetrated the attack.
Confirming the Attack
A minimal amount of photographs of the attack aftermath, all of them unverified, appeared on various social media channels by the beginning of May. They contained gruesome images of a mass of corpses stacked on top of one another or of the local mosque riddled with bullet holes.
The first step to confirm picture authenticity was to use image verification tools such as Photoforensics and Forensically, and then to conduct reverse image searches to ensure they had not been published previously.
With the legitimacy of the photographs confirmed, the next step was to review social media channels and traditional media sources to identify the location of the attack. This was necessary due to the difficulty with geolocating the images based on their metadata.
Information gleaned from Twitter, Facebook and YouTube, coupled with reporting from international news agencies and human rights NGOs, all identified Karma as the site of the massacre.
To truly cover all bases, Google Earth Pro was then utilized to build a map of the area to confirm the geographic claims. Using the conical, pale-green minaret of the mosque and the wall outline from the two above pictures as guides, Karma can officially be confirmed as the location of the massacre.
Three Competing Narratives
At this point, we knew that a massacre occurred, and knew it occurred in Karma. So the question then became: who was responsible for the attack?
We studied the pattern of civilian targeting in Burkina Faso by conducting a general GIS (Geographic Information Systems) investigation through the ACLED database via Tableau. This helped to contextualize the claims of the varying actors, particularly fact-checking the ruling junta’s imprecise assertions.
In the aftermath of the massacre, three principal, competing narratives began gaining traction on social media, which we began investigating:
- The first was that the Burkinabe military forces perpetrated the massacre against unarmed civilians in retribution for an earlier attack by Islamic insurgents on a nearby military installation. This was the narrative reported by international news sources, using credible temporal and geographic references.
- The second narrative, aggressively pushed by the ruling Burkinabe military junta and its aligned media sources, is that it remains unclear who is responsible.
- The third, mostly promulgated on Facebook and YouTube profiles, is that a conspiracy directed from Paris coordinated the attack from a military base in Niger.
Narrative one: The Burkinabe army is guilty
This narrative is supported by the verified images from the massacre aftermath, as well as civil society engagement with survivors. It is further supported by the map which plots potential troop movements from the military base in Ouahigouya to Karma.
Narrative two: “Avoid hasty conclusions”
The second narrative, pushed by the Burkinabe military junta and its affiliated media sources, is that it is nearly impossible to reach definitive conclusions. However, GIS data confutes these claims, and a lack of transparency regarding the investigative commission belatedly organized by the junta undermines this narrative. Finally, a brief overview of jihadist actions in the area showed how a false flag attack perpetrated by insurgents was unlikely.
Narrative three: France staged a false flag attack
The third reactive narrative is one sourced from Facebook and YouTube accounts, which blame Paris for staging a false flag attack. However, a routine RIS on the supposed proof circulating on social media shows that the image reported to be of French soldiers conspiring with Karma locals is several years old and from an entirely different location, and another post actually shows a French military vehicle in Niamey, not Ayorou.
Evidence supporting the first narrative proved most difficult to refute based on the various information collected across intelligence disciplines. Thus, it is possible to confirm it as the most likely, with a high level of confidence.
by Samuele Minelli Zuffa and Christopher Dunn