Date:
Location:
Kings were once thought to hold le droit de glaive, the right to determine who shall be put to the sword. Over time political and juridical developments have greatly diminished that power. However, some years ago already, it was suggested that technological developments had suddenly restored it. At the time, support for this claim was found in the advent of drone warfare and reports that then US President Barak Obama was personally approving lists of names of people to be targeted.
Nowadays, algorithms processing volumes of data far too vast for any human brain independently to sift though, analyze and make sense of are being deployed by state powers (and others) and they now watch over us more closely than ever before. Among other things, systems running such algorithms are capable of detecting patterns hidden within the great bodies of data and meta-data (‘big data’) being collected and to thereby infer conclusions about probable future behavior and occurrences. State authorities are already relying on these predictive capacities to guide their decision-making and actions, inter alia, in the context of crime prevention and sentencing.
The presentation will discuss ramifications that these developments may have on the legal regulation of state conduct in armed conflict. On the one hand it will be suggested that systems processing big data, and certainly more advanced autonomous systems propelled by machine learning and artificial intelligence, may clear some of the proverbial ‘fog of war’ promoting more accurate belligerent action, thus advancing the underlying humanitarian purposes of international humanitarian law (IHL, otherwise known as the laws of war). On the other hand, attention will be drawn to concerns that when applied in armed conflict these systems may prove a double-edged sword, as they are also likely to compromise humanitarian interests. Moreover, from a humanitarian perspective, the harm they cause may significantly outweigh the benefit they yield.
The presentation will not address the—very significant—set of concerns that would arise if the systems under consideration would come under the control of a party that fails to abide by IHL. Instead, it will focus on concerns likely to emerge even when IHL isrespected. Indeed, IHL allows for lethal action and destructive force to be deployed against certain categories of people and certain objects and also for the deprivation of liberty of certain people not as retribution for past deeds, but to prevent a future threat they may pose from materializing. The paper will posit that decisions based on algorithmic analysis of big data threaten to overextend IHL’s preventative logic and unleash more violence than had been anticipated when the law was established. On these grounds it will be suggested that empowering predictive algorithms threatens to further weaken law’s restraining hold on the sword of state power. Indeed the sword, which is acquiring a mind of its own, may strike sooner, more swiftly and more often than it ever has done.
Click on the Picture to Enlarge:
