Saving Lives with Artificial Intelligence

The accurate portrayal of the “fog of war” was gifted to us by Prussian military theorist Carl von Clausewitz. Clausewitz warned that what a commander thinks they know on the battlefield is often wrong, partially wrong, or unknowable in real time due to an accumulation of myriad small errors. The resulting friction is characterized by extreme uncertainty, confusion, and lack of situational awareness experienced during combat.

The instinct to attribute these losses to familiar combat realities — poor coordination, high ops tempo, the fog and friction of war — leaps past one fundamental question: could affordable, modern technology have prevented these outcomes?

Controllers should have waved off merging assets. Soldiers should have known about incoming fire. Planners should have known the targeting threat or soil conditions. Without accurate orientation, decisions and actions were fatally delayed. A digitally twinned battlespace can integrate relevant information, remove errors, and compress the time between observe and orient phases, delivering real time situational awareness that lets warfighters focus on deciding and acting with Artificial Intelligence (AI) by their side.

Technologies exist today to deploy AI-enabled digital twins — closing planning gaps while enhancing situational awareness in real time. A digital twin can be deployed in days or weeks at the cost of one Tomahawk missile, integrating authoritative sources selected by the warfighter, not some contractor back in the States, and tailored for each mission.

A digital twin is not a dashboard or a common operating picture. It is a continuously updated virtual model of the operational environment — reflecting not just asset locations, but the relationships between them, the constraints they operate under, and the conflicts that would arise in certain courses of action. It integrates live Air Tasking Order execution, IFF status, coalition partner activity, electronic warfare conditions, logistics states, and airspace restrictions — all fused in real time. What has been missing is the institutional commitment to deploy it.

The Department of War has recognized the problem. CJADC2 reached a minimum viable capability milestone in early 2024. But incidents continue — because the gap between an experimental result and a fully integrated picture at the tactical edge remains wide. Even as the Department articulates a vendor-agnostic vision, billion-dollar contracts flow to large incumbents for integration work that smaller, more specialized companies already deliver faster and at lower cost. Even worse, incumbents are not incentivized to offer small-company innovation — that would reduce their fees. The Department must implement a deliberate assessment of how fast CJADC2 integration investments result in tangible effects on the battlefield.

Our military fields more sensors and collection systems than any force in history. The problem is what happens after collection. Intelligence, planning, operations, and logistics live in separate systems — reconciled by human analysts under time pressure, if at all. The GAO's 2025 CJADC2 assessment described a "swivel chair model": data manually transferred between disconnected systems to approximate a picture none provides alone. In a peer conflict at the speed of relevance, this model succumbs to the fog and friction of war.

The solution is loose coupling: linking data sources when needed for mission requirements, not vendor preference. Innovative platforms can establish digital twins with federated authoritative sources on demand — drawing from the trusted databases and terrain models simultaneously, on the warfighter's terms.

While AI cannot compensate for broken data architectures, a digital twin foundation with trained AI surfaces issues before they become incidents — including across sensor systems operators might not think to consult. Without a digital twin, AI's value for warfighters is dramatically diminished.

These tragic circumstances have happened before. The same conditions will continue to recur —consider an Indo-Pacific scenario with denser airspace, more capable adversaries, and more severe electronic warfare. It is not just combat: tragic mid-air collisions and ground near-misses at commercial airports persist because we have not twinned the national airspace or trained AI to aid controllers.

The United States does not lack the innovation or human talent to maintain information superiority. We are challenged because we are not deploying the technologies now that integrate what we already know— to allow the hardest fog and friction of war to lift.  Until we commit to digitally twinning a battlespace for the warfighter that is dynamically assessed, continuously updated, and operationally trusted, we will continue to suffer tragic, preventable losses.

The HON Lucian Niemeyer is a former Assistant Secretary of Defense, former White House official, a former staff member of the U.S. Senate Committee on Armed Services, and retired Air Force military engineer.

Timothy Faulkner is a former Senior Defense Executive, former Senior Special Operations civilian in the Department of the Navy, and a retired senior U.S. Army officer.