Lockheed Martin Builds an AI-Powered Counter-Drone Kill Chain in Just 45 Days — And It Works

Lockheed Martin test new AI system intercepts Shahed-style attack drone

The drones reshaping the modern battlefield over Ukraine now have a new American adversary — and the company behind the world’s most advanced fighter jet just proved it can bring one down. In a live-fire test that has drawn significant attention from defence analysts and military planners, Lockheed Martin demonstrated a complete end-to-end counter-drone kill chain, integrating artificial intelligence, battlefield radar, and a battle-proven missile to intercept a Shahed-style kamikaze drone. What makes this test remarkable is not just the outcome — it is the speed at which the entire system was assembled and made operational.


The Kill Chain at a Glance

Before diving deeper, it helps to understand what a “kill chain” actually means in this context. In military terminology, a kill chain is the sequence of steps required to identify, track, target, and destroy a threat — from the moment a radar picks something up to the moment a weapon hits it. Every link in that chain matters. If detection is slow, the target escapes. If targeting data is inaccurate, the missile misses. If the weapon system is not responsive, the window closes.

What Lockheed Martin has built is a tightly integrated version of that chain, specifically engineered for the counter-drone mission. Reporting from Axios confirms that the company successfully intercepted a Shahed-style attack drone using a Joint Air-to-Ground Missile (JAGM) fired from what it calls the Grizzly containerised launcher. This marks the first time Lockheed has ever employed JAGM to destroy an unmanned aerial threat of this type, and the test validated the complete system working as a single, connected whole.

At the heart of it all is Sanctum — Lockheed’s AI-enabled battle management software — which received targeting data from Fortem Technologies’ R40 radar, processed the track, and cued the weapon to fire. From first radar contact to missile impact, the system worked seamlessly.


The Threat It Was Built to Counter: The Shahed and Its Legacy

To appreciate why this test matters, you need to understand the weapon it was designed to kill.

The Shahed-136 is an Iranian-designed loitering munition — a one-way kamikaze drone that flies to its target and detonates on impact. Russia has manufactured it domestically under the designation Geran-2 and has deployed it in their hundreds in mass overnight strikes against Ukrainian cities and power infrastructure. These attacks, which have become a grim fixture of the conflict since 2022, have repeatedly knocked out electricity for millions of civilians during winter months and stretched Ukraine’s air defence resources to their limits.

The drone’s effectiveness stems from a combination of characteristics that make it simultaneously cheap to produce and genuinely difficult to defeat. At an estimated cost of between $40,000 and $80,000 per unit, it is orders of magnitude cheaper than the surface-to-air missiles used to intercept it. Its small airframe gives it a low radar cross-section, making it harder to detect and track at range. It flies at low altitude, hugging terrain to avoid radar horizons, and its slow, steady approach can be masked amid electronic noise. When dozens are launched simultaneously from different directions — a common Russian tactic — they create a saturation problem that even sophisticated integrated air defence systems struggle to solve.

Ukraine has been forced to expend expensive interceptors like the Patriot’s PAC-2/PAC-3 missiles and IRIS-T rounds against threats that cost a small fraction of what it takes to shoot them down. The economic mathematics are deeply unfavourable for the defender. A Patriot interceptor costs roughly $2 to $4 million. A Geran-2 costs perhaps $50,000. When Russia launches 50 drones in a single night, and Ukraine burns through interceptors trying to stop them, the attacker wins the economic exchange even if every drone is destroyed.

This cost-exchange problem has been the central challenge facing not just Ukraine, but Western defence planners who have been watching the conflict as a preview of what future high-intensity warfare could look like. The question of how to defend against mass drone attacks without bankrupting yourself in the process has become one of the most pressing problems in contemporary air defence.


Sanctum: The AI Brain Orchestrating the System

The component that sets this kill chain apart from previous counter-drone solutions is Sanctum, Lockheed’s AI-enabled battle management system, which functions as the central software intelligence of the entire architecture.

In a modern air defence scenario — particularly against a mass drone attack — the cognitive burden on human operators is immense. Dozens of low-flying threats may be inbound simultaneously, approaching from multiple vectors, at varying altitudes, on routes designed to exploit gaps in radar coverage. Operators must correlate tracks from multiple sensors, assess threat priorities, allocate available weapons, and issue fire commands — all within a timeframe measured in seconds, not minutes. The human brain, however well-trained, has limits that a sufficiently complex attack can overwhelm.

Sanctum is designed to address exactly that problem. It ingests data from multiple sensor types simultaneously, correlates tracks across sensor feeds, assesses threat priority using AI-driven analysis, and assigns available weapons automatically. This does not eliminate the human from the loop — a human operator retains the authority to approve or override engagements — but it dramatically reduces the cognitive load, allowing a small crew to manage a high volume of simultaneous threats that would otherwise exceed their capacity to handle.

In the context of the test, Sanctum received targeting data from the R40 radar, maintained a continuous track on the Shahed-style target, and cued the JAGM launcher at the appropriate moment. The system handled the targeting solution automatically, meaning the time between detection and weapon release was compressed to well below what a manually operated system could achieve.

This kind of AI-curated, software-defined air defence is where Lockheed sees the broader trajectory of the counter-drone mission heading — a recognition that the volume and variety of future drone threats will simply exceed the capacity of human operators working with conventional tools.


Fortem Technologies’ R40 Radar: Eyes Built for the Mission

One of the less-publicised but critically important elements of this kill chain is the radar providing the targeting data that feeds Sanctum. Lockheed integrated Fortem Technologies’ R40 — a truck-portable detection and tracking sensor developed specifically for counter-drone operations.

This matters more than it might initially appear. Many existing air defence radars were designed primarily to track aircraft, helicopters, and missiles — large, fast-moving targets with substantial radar cross-sections. Adapting those systems to detect small, slow, low-flying drones — which are precisely the kind of targets optimised to avoid conventional radar — introduces performance compromises. The radar is effectively being asked to do something it was not optimised for.

Fortem Technologies, a Utah-based counter-drone specialist that has built a strong reputation in the UAS detection and defeat space, developed the R40 from the ground up to find exactly the kind of targets the Sanctum-JAGM system is designed to kill. Its sensor signatures and tracking algorithms are tuned to the flight profiles of small and medium unmanned aircraft, making it far more capable in that specific role than a general-purpose system pressed into counter-UAS duty.

The fact that the R40 is truck-portable is equally significant. It can be driven to wherever the threat is, set up quickly, and integrated into the kill chain without requiring permanent installations or specialised infrastructure. That mobility characteristic mirrors the broader design philosophy of the system as a whole.


The Grizzly Launcher: Mobility as a Force Multiplier

Perhaps the most operationally relevant aspect of the entire system — at least from the perspective of military commanders who have to actually deploy and protect it — is the Grizzly containerised launcher that fires the JAGM.

Rather than a fixed installation requiring concrete emplacements, utility connections, and permanent infrastructure, the Grizzly is packaged inside a standard military shipping container. That single design decision has profound implications for how the system can be used. It can be loaded onto a flatbed truck, a railcar, or a cargo aircraft. It can be positioned at a forward operating base, moved when the tactical situation changes, and repositioned quickly if the original site comes under threat.

This matters enormously in contested environments. Fixed air defence sites are high-priority targets. The moment an adversary identifies the location of a Patriot battery or an equivalent system, it becomes the target of the next strike. Static systems require elaborate protection, hardening, and redundancy precisely because they cannot move. A containerised launcher that can be repositioned within hours fundamentally changes the threat calculus for an adversary trying to suppress your air defences.

The Grizzly concept also speaks directly to the logistical realities of rapid deployment scenarios. If a crisis emerges and forces need to establish a defensive perimeter quickly — around a port, an airfield, a critical piece of infrastructure — the ability to truck in a complete air defence capability and make it operational in hours rather than days is not a luxury. It is a basic operational requirement.


JAGM: A Battle-Proven Missile Finds a New Purpose

The weapon fired from the Grizzly launcher is not a new development. The Joint Air-to-Ground Missile has been in service with U.S. Army and Marine Corps aviation units for several years, having been developed originally as the replacement for the iconic Hellfire missile on attack helicopters and unmanned platforms.

JAGM uses a multi-mode seeker that can track and engage targets via semi-active laser guidance, millimeter-wave radar guidance, or a combination of both simultaneously. This flexibility makes it effective across a wide range of conditions — through smoke, dust, bad weather, and against targets that may not have a strong laser return. Its shaped-charge warhead is capable of defeating armoured vehicles, bunkers, and fortified positions, and it carries effective range out to roughly 16 kilometres (approximately 10 miles).

What this test confirmed is that JAGM’s tracking and guidance systems are capable of engaging a fast-moving aerial target — not just the ground vehicles and static targets it was originally designed for. Repurposing an existing, mature missile for the counter-UAS mission represents a significant economy of investment. Rather than developing and funding a brand-new interceptor from scratch — a process that typically takes years and costs billions — Lockheed is drawing on a weapon that is already in full production, already in the supply chain, and already thoroughly understood by the forces that operate it.

This approach also accelerates the path to fielding. There are no new manufacturing lines to stand up, no new maintenance training to develop, no new logistics chains to establish. The missile is already there. The innovation lies in integrating it into a new architecture and confirming that it can do a new job.


45 Days: The Number That Is Disrupting a Slow Industry

If there is a single figure from this announcement that has resonated most loudly across the defence community, it is the timeline.

Lockheed Chairman, President, and CEO Jim Taiclet stated that the company assembled the complete counter-drone capability — integrating the radar, AI battle manager, missile, and launcher into a working end-to-end system — in under 45 days. Lockheed Vice President and former Trump White House official Jalen D. put it plainly in a widely-circulated social media post: “A complete counter-drone kill chain delivered in 45 days. That’s the speed today’s threats demand.”

In the context of normal defence procurement timelines, 45 days is essentially instantaneous. Major weapons programs — aircraft, ships, missile systems — routinely take five, ten, or fifteen years from initial concept to fielded capability. Even relatively straightforward upgrades to existing systems can take years to work through requirements definition, contracting, development, testing, and approval processes. The institutional machinery of defence acquisition is, by design and by historical habit, a slow-moving engine.

The speed of the Grizzly-JAGM-Sanctum integration was possible precisely because all three components already existed. Lockheed did not invent a new radar, a new missile, or a new AI system. It took proven, mature technologies and integrated them into a coherent kill chain — a task that the company says took less than seven weeks from start to live-fire confirmation. That is the kind of timeline that an adversary deploying Shaheds by the thousands cannot easily outpace.

The broader implication is a model for how the defence industry might respond to rapidly evolving threats. Rather than waiting for a clean-sheet development program to work its way through the system over a decade, the answer may often lie in creative integration of what already exists — faster, cheaper, and immediately actionable.


The Wider Context: Closing a Critical Gap in Western Air Defence

Lockheed’s test does not exist in isolation. It arrives against a backdrop of sustained institutional concern within the U.S. military and its allies about the widening gap between the volume of drone threats anticipated in a near-peer conflict and the capacity of existing air defence systems to handle them.

Military commanders and Department of Defence officials have repeatedly flagged this mismatch in public testimony, strategy documents, and capability reviews. Programs like the Indirect Fire Protection Capability (IFPC) — which combines radar, directed energy weapons, and interceptors on a single integrated platform — reflect the urgency with which the Pentagon is pursuing solutions. The IFPC effort acknowledges that point-defence interceptors alone cannot solve the mass drone problem, and that a layered approach combining kinetic and non-kinetic effects is needed.

Directed energy weapons — high-powered lasers and high-power microwave systems — offer the theoretical promise of near-zero marginal cost per kill, which would transform the cost-exchange problem. But these technologies are still maturing, with limited range, power challenges, and atmospheric performance variability that constrains their reliability in all conditions. Until directed energy matures to true operational readiness at scale, kinetic interceptors remain the primary answer — and the economics of that answer depend heavily on whether the interceptor costs $2 million or $50,000.

JAGM, at a unit cost substantially lower than premium surface-to-air interceptors, represents a step toward better cost economics for the defender. It is not the final answer, but it is a more favourable exchange ratio than firing a Patriot round at a $50,000 drone.


From Live-Fire Test to Fielded Capability: What Comes Next

It is important to be clear about what Lockheed’s test does and does not prove. A successful live-fire intercept of a single Shahed-style drone under controlled test conditions is a meaningful and encouraging data point. It is not a fielded operational capability.

The path from test to deployment typically involves additional trials — stress testing the system under more challenging scenarios, evaluating performance against jamming and electronic countermeasures, validating reliability across varied environments and weather conditions, and verifying that human operators can actually manage the system effectively under the time pressure of real operations. Then comes the acquisition process: requirements validation, programme of record decisions, contracts, production, training, and logistics. That process, even under favourable conditions, takes time.

What the test does demonstrate is that the architecture works — that AI-enabled battle management, dedicated counter-drone radar, mobile launcher, and proven missile can be integrated into a coherent kill chain and made to function as intended. For a defence community that has been debating what counter-drone capability should look like, that proof of concept carries real weight. The argument for funding and developing this capability is now grounded in a concrete, demonstrated result rather than a theoretical proposal.

Whether Sanctum and the Grizzly-JAGM combination will find a place in formal U.S. or allied acquisition programmes remains to be seen. But the conversation about how they might be used has certainly shifted.


Final Thoughts

The drone threat that has defined the air war over Ukraine is not going away. If anything, it is proliferating. Adversaries across the globe have observed the effectiveness of cheap, mass-produced loitering munitions against sophisticated air defences and have drawn the obvious conclusions. The next major conflict — wherever and whenever it occurs — is increasingly likely to begin with drone swarms before a single soldier or warship enters the contested space.

What Lockheed Martin has demonstrated is that a credible answer to that threat can be built from existing technology, integrated in weeks, and made operational in a mobile package that can go where the force goes. The combination of AI-driven targeting, purpose-built detection radar, a battle-proven missile, and a containerised launcher is not a perfect solution to the drone problem — no single system is. But it is a working solution, and in an environment where adversaries are buying Shahed clones by the thousand, working solutions delivered in 45 days are exactly the kind of answer that matters.

The Shahed changed the battlefield. The question now is whether Western industry can change fast enough to keep pace — and Lockheed’s latest test suggests the answer might be yes.