In 2026, the history of modern warfare reached a point of no return. The United States moved from experimenting with artificial intelligence to its systematic integration into military planning, intelligence, and combat operations. The logic of warfare itself is changing: the advantage increasingly belongs not to the side with more missiles, aircraft, or drones, but to the side that can transform data into decisions faster.

United States: From Experiment to Combat System

One of the key instruments of this transformation is Palantir’s Maven Smart System. The system integrates data from multiple sensors, helps identify and prioritize targets, and supports the selection of appropriate means of engagement.

The scale is no longer measured by laboratory tests. During the U.S. operation against Iran, which began on February 28, 2026, American forces struck more than 13,000 targets during the first 38 days. These included:

• more than 2,000 command-and-control facilities;
• 1,500 air-defense facilities;
• 1,450 industrial facilities.

According to the Arms Control Association, Maven was used to identify priority targets and support weapons-selection decisions.

This demonstrates the fundamental shift: AI is gradually taking over the most labor-intensive part of the targeting cycle, processing enormous volumes of intelligence data. Previously, a headquarters might spend hours correlating satellite imagery, UAV data, radar information, and other sources. An algorithmic system can perform this task in near real time.

From Decision Speed to the Speed of War

In conventional warfare, superiority was determined by the number of forces, weapons range, and the quality of intelligence. A new parameter has now been added: the speed at which the sensor-to-intelligence cycle is completed.

Conceptually, the chain looks like this:

sensor → data → analysis → decision → weapon → result.

The less time this cycle takes, the harder it becomes for an adversary to change position, relocate a command post, activate air defenses, or move a target out of the engagement zone.

This is why AI is becoming not merely an auxiliary technology, but an element of combat power.

The principle is already moving beyond ground-based systems. In April 2026, the U.S. X-62 VISTA demonstrated during testing that AI agents could receive data from an infrared sensor and autonomously control the aircraft to intercept an airborne target in real time, according to the Air Force Test Center.

The issue is no longer simply whether AI can “advise” a pilot or operator. An algorithm can directly generate commands for the platform.

A New Factor of Strategic Deterrence

This creates a challenge for the traditional system of strategic deterrence. Doctrines based on the assumption that an adversary will have time to assess a situation, negotiate, and make a decision are increasingly confronted with systems operating on a scale of seconds and minutes.

The combination of AI with autonomous platforms, cyber weapons, long-range missiles, and space-based intelligence systems is particularly dangerous.

The future balance of power will therefore be determined not only by the number of warheads, but also by a state’s ability to see the battlefield, predict an adversary’s actions, and close the engagement loop faster than its opponent.

Ukraine: The War Has Already Created Its Own AI Ecosystem

For Ukraine, this is not a theoretical debate. The full-scale war has generated a unique body of real-world combat data that has become one of the most valuable resources for developing military AI.

On January 20, 2026, Ukraine, together with Palantir, launched Brave1 Dataroom, a secure environment for testing and training military AI models using real battlefield data. The first focus was autonomous detection and interception of aerial threats, including Shahed-type UAVs, according to Ukraine’s Ministry of Defense.

Within just a few months, the scale had changed significantly:

• more than 100 Ukrainian companies gained access to Dataroom and used real structured datasets to train and validate models;
• more than 80 models were being developed and tested, including systems for detecting and intercepting aerial targets, according to the Ministry of Defense of Ukraine.

This is fundamentally important. Ukraine possesses something that cannot simply be purchased: real data from modern warfare.

These datasets include visual and thermal characteristics of aerial targets collected under different weather conditions, at different times of day, and using different types of sensors.

In May 2026, Reuters reported that Ukraine was already using AI-based solutions to analyze aerial attacks, process large volumes of intelligence, and plan operations deep inside adversary territory.

The Next Stage: “Warfare of Operating Systems”

Ukraine is already moving from individual AI-based solutions toward an integrated combat environment.

In March 2026, the AI Defence Centre was established with the task of integrating AI into military command and control. The concept envisages combining data from different systems across the entire 1,200-kilometer frontline and creating a unified information environment for decision-making, according to Reuters.

This represents the next stage of warfare: a competition not between individual drones or missiles, but between combat operating systems that integrate intelligence, weapons, command and control, and algorithms into a single loop.

The adversary is adapting as well.

russia is developing algorithms for automatic target recognition, UAV tracking, navigation under electronic-warfare conditions, and the use of AI in cyber operations.

The weakness of the adversary lies in fragmentation and dependence on individual developments. However, its advantage in the scale of available resources means that the technological gap could narrow rapidly.

A Roadmap for Ukraine

To neutralize the adversary’s advantage and ensure long-term technological resilience, Ukraine needs to move from individual AI projects toward a national military AI system:

1. AI-assisted operational planning. Analytical systems should combine intelligence, logistics, satellite imagery, UAV data, and the adversary’s electronic footprint to rapidly model operations, including precision actions by special forces against artillery, missile systems, and logistics.

2. Autonomous “last mile.” Ukraine should develop computer vision and autonomous terminal guidance capable of operating without continuous communication with an operator and withstanding electronic-warfare interference.

3. Sovereign data infrastructure. Combat experience must be systematically converted into structured data that can be continuously used to train Ukrainian models.

4. Systematic scaling. One hundred companies and 80 models represent an important starting point, but achieving systemic superiority requires scaling from hundreds of developers to thousands and from dozens of models to a full national ecosystem.

The war in Ukraine has become one of the world’s principal testing grounds for next-generation technologies. Ukraine possesses a unique advantage: real-world combat experience, an enormous body of data, and the ability to rapidly test technologies under conditions of high-intensity warfare.

But this is not enough.

The next technological race will not be determined by who has the better model, but by who can transform data into decisions faster, decisions into action, and combat experience into the next version of an algorithm.

Artificial intelligence is ceasing to be merely a decision-support tool. It is becoming an independent factor of strategic advantage and one of the principal criteria of a state’s military power.