The United States Army has introduced "deep sensing" as a feature of its battlefield strategy, integrating high-resolution sensors and AI to track and anticipate enemy movements.
At the centre of these developments are platforms like HADES and TITAN, developed by the US Army in collaboration with large corporate consortiums. These systems are designed to collect, analyse, and fuse data from airborne sensors, satellites and ground networks to create a real-time, multi-domain picture of the battlefield.
In a new blog post for international law blog OpinioJuris, researchers Klaudia Klonowska (Asser Institute) and Sofie van der Maarel (Netherlands Defense Academy) argue that the seemingly neutral and technical term “deep sensing” is echoing familiar narratives of enhanced precision, speed, and scalability seen in other AI-driven warfare technologies.
Profound transformation
Klonowska and Van der Maarel suggest that the use of the "deep sensing" label is strategic in that it disguises the intricate ethical, legal, and societal implications of the technology underpinned by surveillance and large data analysis.
The authors argue that “deep sensing” is not simply about gathering more data, but about how that data is turned into decisions on the battlefield. Platforms like HADES and TITAN produce datasets so large that humans alone cannot extract any meaning from it. Thus, 'deep sensing’ technologies rely on algorithmic systems to actively interpret it, prioritise it, and convert it into actionable information for commanders and autonomous systems.
What concerns the authors is how the term “sensing” suggests passive observation, yet the underlying systems are actively shaping threat perceptions of military actors. By focusing on data collection, defense contractors and officials downplay the growing role of algorithms in targeting and threat assessment, according to the authors.
