How Data Drives the Militarisation of Tech

There’s a revolution occurring in how war and conflict are waged. New data-intensive systems are being developed; battlefields are testing grounds for a new generation of tech; and commercial tech infrastructure is now supporting military operations.

Data plays a key role in this revolution. Data is used to train and test systems, and the systems are fed data to target operations, communities, and individuals.

While intelligence has long informed warfare, now we’re seeing the very same dynamics that gave rise to surveillance capitalism feed a new era of innovation, what we call the Militarisation of Tech. Similar tools, investment vehicles and even the same firms are extending development to warzones, and bringing the work back to home.

Understanding the role data plays in these systems is crucial to grasping their impact. Militarisation of Tech is affecting everything from everyday technologies to high-stakes contexts such as warfare. This is the next great challenge for the protection of privacy, autonomy and dignity.

Data Revolution in War

1. Adapting tools from commercial world

Surveillance capitalism devised expansive methods to target people based on data, using analytics and statistical methods.

Many of the Big Tech firms that developed these techniques also developed cloud services. Governments are moving their health, welfare and even intelligence data into these cloud services rather than to run their own servers. Using Big Tech’s cloud services, governments are now able to process this data at a previously unimaginable scale. This also creates new dependencies as governments are now reliant on these firms to manage and maintain more and more of the functions of the state.

With the rush to develop AI models, these firms with models are now keen to sell their services to the military, to find ways to help military organisations to manage their data as well.

These cloud services now offer AI tools to help manage and understand the data held by companies and governments.

Meanwhile, Defence Tech firms have emerged to generate and harness data, and apply data analytics to more novel and advanced applications in war. These firms are also looking to broaden markets, and have also developed systems to deploy them in policing and at borders.

2. Learning from Data: Building Systems

While intelligence has long informed military action, now intelligence can feed the development of systems. The data coming from conflict zones and warfare, gathered through surveillance including aerial surveillance, hacking, device exploitation, and other forms of mass data surveillance, can be used to train systems.

For instance, by accessing the database of a national ID system of an adversary or a social media platform used by a targeted community, a facial recognition system of a civilian population can be developed.

In another instance, all the data held by a military operation can be used to generate a Large Language Model (LLM) and receive natural-language queries, to assist in the management and waging of war.

LLMs can now be trained on the languages and contexts of adversaries. Through the accumulation of both written and then intercepted voice communications, they can now be used to understand dialects, gauge public reaction, influence social media discussions, and more.

3. Data-Led Targeting

Actionable insights is a key objective of intelligence. Now through advanced processing of vast volumes of data the production of actionable insights vastly increases, even in real time. An actionable insight can now includes processed and contextualised data, as well as recommended actions based on them.

For instance, vast amounts of data on telecommunications activity, such as connections, movements, interactions, and content that was intercepted could be accumulated to conduct surveillance by developing tools to analyse datasets.

Now in order to produce actionable insights, data-intensive systems can be built to handle and process vast amounts of diverse data by design, incorporate disparate data types and in real-time. If you think of big-data as the raw materials then data-intensive-systems would be the factories that turn those raw materials into finished useable products – or in this case actionable insights.

Combining facial recognition systems with text messages, social media posts and other telecommunications data, the challenging process of generating targets can be automated and sped up. Drones can be assigned people to recognise and track. And these actionable insights can be used to target weapons.

Why is this a problem

1. Fusion of civilian and military data

One of the main problems with the development of data-intensive systems for military and defence in parallel to civilian and consumer contexts is that, by definition, none of these systems can function without the massive surveillance infrastructures that support their functionality and their constant demand for data, be it satellite data, social media data, telecommunications data or other.

As an example take the revelations that U.S. military regularly purchased granular location data harvested from seemingly innocuous consumer apps — like a Muslim prayer app with over 98 million downloads — via data brokers who had amassed this data seemingly for marketing purposes but in the end for much more problematic uses.

This hunger for large sets of diverse data creates a structural dependency on continuous, large-scale monitoring of individuals, communications, networks, environments, and behaviours - blurring the line between combat zones and everyday civilian life and vice versa. This incites ever more mass surveillance practices.

2. Fusion of civilian and military infrastructures

As such systems are further integrated into existing technologies and public infrastructures, such as smart cities, the scope of surveillance extends beyond military targets, enveloping entire populations under the guise of ‘national security’, often blending data sources and insfrastructures across civilian and military contexts.

One example that shows the dual-use nature of smart city infrastructure is Belgrade’s Safe City project in Serbia. This system has over 1000 smart cameras equipped with facial and license plate recognition, initially deployed with the purpose of improving public safety and manage traffic, but able to in parallel support law enforcement and national security operations. The agreggation of data from multiple public and widespread sensors and monitoring systems enables real-time tracking of individuals and plainly illustrates the issue of lending civilian infrastructure to military forces. In another, Israel’s National Cyber Directorate has urged the disconnection of home internet-connected security cameras because of the concern that Iran may be exploiting them for intelligence and targeting missiles. Russia similarly warned civilians regarding cameras and the use of dating apps and social media by residents near the border of Ukraine.

3. Normalising securitisation and militarisation

These ubiquitous integrations risk the normalisation of surveillance practices and security driven policies. As military systems seep into law enforcement, border control, public health, and even education, the distinction between national defence and domestic governance begins to collapse. This ‘securitised state’ no longer responds only to identifiable threats, but instead operates through an invisible logic of permanent suspicion, where every human is visible, traceable, and potentially dangerous. This leads to more data being generated on all people.

This dynamic is further reinforced by a global trend of steep escalation in defence budgets. In 2024 alone, global defence spending reached a record $2.72 trillion, a 9.4% increase from the previous year and the highest annual jump since the Cold War. Governments, venture capitalist, sovereign investment funds, and companies across the globe are heavily investing in AI-powered surveillance, cyber capabilities, and autonomous systems – many of which are reliant on the vast processing of people’s data.

With defence spending surging globally and the normalisation of surveillance spreading across sectors, the militarisation of society risks becoming both permanent and invisible in the form of the Militarisation of Tech.

The Need to Understand the Data-Intensive Systems

What begins as defence quickly becomes an all-seeing mode of governance, one that treats entire populations as data sources and potential threats.

The dominant firms of tomorrow are using tools developed and sold for waging war and conflict. They will likely be the companies who will run our domestic social programmes too. People’s data continues to feed their dominance.

Confronting this reality requires a critical understanding of how data-driven systems function, how they process people’s data, and what they ultimately cost, not just in terms of data, but also in terms of democratic integrity, civil liberties, and the right to live beyond the shadow of constant surveillance.