The US Army is trialing an advanced AI system, ‘Scylla’, touted to detect threats from a significant distance without requiring new hardware. Named after the fabled Greek sea monster, Scylla has been under evaluation at the Blue Grass Army Depot, Kentucky for eight months. This munitions depot, previously a chemical weapon storage site, is seeing Scylla enhance its security infrastructure. Chris Willoughby, overseeing electronic security at BGAD, praises Scylla’s high threat detection accuracy, reducing false alerts caused by environmental factors.
The Physical Security Enterprise and Analysis Group (PSEAG), managing the Scylla trials, has optimized the AI to analyze individuals’ features, behaviors, and potential armaments in real-time, minimizing unnecessary security responses. According to Drew Walter, Deputy Assistant Secretary of Defense for Nuclear Matters, Scylla utilizes existing video feeds to quickly assess situations, aiding the PSEAG’s mission to protect national strategic assets.
The system’s real power lies in its ability to monitor large areas using drones and cameras, providing more precise threat identification compared to humans. It can reportedly recognize a weapon from a thousand feet using multiple camera feeds, as demonstrated in simulated tests at BGAD.
The Navy and Marine Corps plan to conduct their own Scylla tests at Joint Base Charleston in the near future. Meanwhile, the Department of Defense highlights Scylla’s cost-effectiveness and adaptability, though it’s unclear if the military employs a standard commercial version or one personalized for its specific needs.
While Scylla’s commercial arm asserts its AI is unbiased and commits only to ethical applications, its activities in regions with contentious human rights records have raised questions about these ethical promises.