US researchers have created a new algorithm that gives us deeper insight into how the Islamic State (IS) execute their military strategy.
Paulo Shakarian of Arizona State University, who co-authored a study, identified how the militia altered its activities with opposition air strikes as well as explored reasons for the group's increased use of improvised explosive devices (IEDs).
One of the most important conclusions of the paper points out that following airstrikes, IS resort to using more IEDs.
Noel Sharkey, a professor of artificial intelligence and robotics at the University of Sheffield, told The Huffington Post that the research now confirms a causal relationship between both factors -- that is to say, the airstrikes are causing a change in militant tactics.
"One of the main finding was that following airstrikes, ISIS moves to attacks using Improvised Explosive Devices," he explained.
"This would appear fairly obvious from an examination of the data. However, this system moves from the notion to correlations in the data to a causal relation."
While it is safe to say that the algorithm is a form of artificial intelligence, Sharkey also warns against using the term too lightly as it's not a learning system.
"What we have here is a useful technique for analysing military strategy using a new algorithm.
"The system does not work on the fly to work out the ISIS strategy.
"A lot of data first has to be collected and then rules are extracted from it. It is not an adaptive algorithm that can adjust it predictions on the basis of incoming data."
However, the algorithm does give analysts the freedom to have a definite set of reasons why certain patterns of IS activity occur.
For example, why did IS increase their arrests following Syrian airstrikes?
Shakarian told the BBC, that it may have been an attempt to flush out Syrian intelligence agents -- a conclusion made possible by the algorithm.
The ability to identify causal relationships in the fight against IS is undoubtedly a useful tool.
Sharkey does argue that in order for artificial intelligence to become an invaluable part of the war on the ground, scientists will need to create an algorithm that will work "on the fly," adapting to changing environments.
"I imagine that what a military commander needs is an AI system that can take many contextual factors into account in the field and with sufficient data about the enemy commander to predict how they will operate in these conditions."