More than 100K personnel use Maven Smart System: Pentagon official

DefenseScoop | Jon Harper

The Pentagon has expanded its user base for Palantir’s Maven Smart System to over 100,000 personnel following operational demands during Operation Epic Fury. The AI-powered software fused disparate data streams and targeting workflows, enabling U.S. forces to strike 13,000 targets across a 38-day campaign. To solidify this capability, Deputy Defense Secretary Steve Feinberg directed the transition of the platform into a formal program of record by fiscal year-end, following a contract ceiling increase past $1 billion.

However, expanding AI capabilities across combatant commands has exposed critical hardware constraints. Severe compute bottlenecks remain. Consequently, Deputy Undersecretary James Mazol is petitioning Congress to fund the AI Arsenal initiative for secure, government-owned sovereign data centers. Simultaneously, Chief Digital and AI Officer Cameron Stanley emphasized leveraging data architecture to streamline end-to-end logistics, aiming to better prepare military supply chains for high-intensity, multi-month conflicts.

Comment

The rapid scaling of the Maven Smart System during Operation Epic Fury demonstrates a structural shift from network-centric targeting to processing-bounded warfare. While automated target recognition accelerates kill chains, algorithmic throughput faces strict operational limits imposed by classified hardware availability rather than raw data volume. Sovereign compute infrastructure, such as the proposed AI Arsenal, reflects the reality that advanced machine learning models cannot operate on unclassified commercial cloud networks during high-intensity conflict.

This computing bottleneck mirrors the early deployment of the Semi-Automatic Ground Environment air defence network during the Cold War, where processing capacity across IBM AN/FSQ-7 installations bounded operational throughput. Then as now, scaling real-time command capabilities like SAGE or the Maven Smart System depends far more on forward-deployed processing hardware than on software refinement alone.

Strategic Question for Discussion

If theater-level processing hardware like the proposed AI Arsenal cannot keep pace with target generation in high-tempo engagements, which constraint will prove more severe for the Maven Smart System: cloud-to-edge bandwidth bottlenecks or physical munition stockpiles?

The operational pattern from Operation Epic Fury suggests that physical munition exhaustion will constrain targeting cycles before bandwidth limitations disrupt edge processing. While containerized compute nodes can temporarily cushion bandwidth degradation, sustaining 13,000 strikes over 38 days rapidly depletes precise guidance kits and air-to-surface inventories. Consequently, algorithmic targeting speed ultimately shifts the primary operational failure point from software processing to industrial production lead times.

Share your assessment in the comments below.

To read in details click here