Journal Article: Algorithmic innovation beyond the boundaries of the law

On March 20, 2020, one week after the Dutch government announced the first national lockdown of the COVID-19 pandemic, a group of military data scientists, programmers, and behavioral analysts met on a military base in the center of the Netherlands. They were welcomed by their new commander, who instructed them on their task: to collect and assess all relevant online information and disinformation relating to the pandemic. Using machine learning and predictive modeling, their goal was to turn the data into analyses informing military and civilian decision-making during the unfolding crisis. The commander informed the group that their work charted new territory. “Prepare yourself for experiments”, he said, because “we are reinventing ourselves”.

What was presented as a proactive response to a pressing national crisis, was in fact an unsolicited, highly problematic initiative with no legal basis. Despite warnings from the MoD’s in-house military lawyers, however, the newly formed team – baptized the ‘Land Information Manoeuvre Centre’ (LIMC) – continued its contested activities. Not until eight months later, after Dutch investigative journalists from the NRC daily newspaper published about the LIMC and its unlawful practices, did the unit cease its activities.

Following the NRC investigation, the Dutch equivalent of a Freedom of Information request (‘Woo-verzoek’) resulted in the publication of a collection of more than 400 military orders, minutes, and email traffic between LIMC commanders, MoD lawyers, and high-ranking military officers. In their latest academic publication in Global Society, IRW’s Dr. Lauren Gould and Jack Davies team up with Dr. Marijn Hoijtink and Martine Jaarsma to critically analyse this unique data collection.

Drawing on ‘assemblage thinking’, Gould, Davies, Hoijtink, and Jaarsma trace, through the LIMC case, how algorithmic innovation in contemporary warfare unfolds. In doing so, their analysis provides a unique and detailed insight into how the Dutch military develops and assembles algorithmic technology in the face of internal and external contestation, and how those contestations are in turn negotiated and reassembled in the face of an overarching and highly performative rhetoric around the need to experiment, engage, and lead in “information manoeuvre”. They find that the LIMC team was able to ward off critique and hold the centre together by embracing practices of entrepreneurialism, informality, and experimentation.

In their article, Gould, Davies, Hoijtink, and Jaarsma show that the alliances and contestations among civil and military actors that they uncover are vital for the ways in which algorithmic capabilities in warfare are actually innovated, diffused, and put to use in today’s theatres of war. Finally, they argue that the practices of entrepreneurialism, informality, and experimentation seen in the LIMC case have important political repercussions beyond the Dutch case, as they push the expansion of military surveillance, pattern-finding, and targeting while undermining the rule of law and democratic accountability.

Find the full open-access article here.