The Age of Mass Intelligence: What America Is Learning From Ukraine, AI and the New Battlespace

At the Intelligence & National Security Summit, the most important question was not whether artificial intelligence will replace the analyst. It was whether the United States can reorganize itself quickly enough for a world in which intelligence, weapons and decisions are produced at unprecedented scale.

By Ashley Greer

WASHINGTON — For two days at the Intelligence & National Security Summit, senior intelligence officials, military leaders and technology executives talked about artificial intelligence.

But by the end of the conference, AI seemed almost secondary.

Again and again, the conversation returned to the same underlying problem: scale.

There is too much data for human beings to examine. Too many sensors producing it. Too many systems that need to communicate with one another. Too many inexpensive drones that can be manufactured and lost. Too many satellites potentially entering orbit. And too little time between discovering something and needing to act upon it.

The intelligence problem is no longer simply obtaining information.

Increasingly, it is constructing a system capable of finding the important signal inside an almost limitless volume of information, determining whether that information can be trusted, connecting it with other sources and turning it into action—sometimes within seconds.

That changes the role of artificial intelligence.

AI is not merely another analytical tool being introduced into the intelligence community. It is becoming part of the architecture required to operate an information environment whose scale is beginning to exceed unaided human cognition.

And Ukraine is demonstrating what the military version of that transformation may look like.

From scarcity to abundance

For most of intelligence history, collection was extraordinarily difficult.

Satellites were rare and expensive. Surveillance platforms were specialized. Analysts worked with information acquired through systems that governments spent enormous amounts of money developing and protecting.

That world is disappearing.

Commercial satellites, unmanned systems, open-source intelligence, ubiquitous sensors and enormous databases are creating an environment in which the problem can become abundance rather than scarcity.

One presentation described a future potentially containing thousands upon thousands of commercial satellite constellations. Another discussion focused on the difficulty of keeping pace with the sheer size of available data.

The recurring question became remarkably simple:

What should a human actually look at?

That is where AI increasingly enters the intelligence process.

Officials described systems capable of continuously processing information, triaging it and tipping analysts toward material requiring human attention. In one NSA example recorded in my notes, AI helped process approximately 7,500 maps in seven months, work that speakers said previously could have taken years.

The significance is not that the machine replaced the analyst.

It changed what reached the analyst.

That distinction matters.

When an algorithm determines which images, signals, anomalies or relationships deserve human attention, AI has already influenced the analytical environment before a person makes a judgment.

The human may remain “in the loop,” while the machine increasingly determines what appears inside the loop.

The new intelligence bottleneck

That makes another word repeatedly heard throughout the summit especially important:

integration.

The intelligence community has spent decades building extraordinarily capable individual systems. The emerging challenge is getting them to work together.

Data must first be discoverable. Then accessible. Then understandable. Then trusted.

Only afterward can it become useful.

The problem was summarized repeatedly in different forms: information that arrives too late has limited value, and information whose provenance cannot be established may be equally useless.

Speed therefore cannot simply mean producing an answer faster.

The system must also establish where information came from, how it was generated, whether it has been altered and how much confidence should be placed in it.

That produces a tension running through the AI revolution.

The battlespace demands speed while intelligence demands trust.

Solving both simultaneously requires something more complicated than installing better algorithms. It requires connecting databases, sensors, agencies, commercial systems and military platforms that were not necessarily designed to communicate with one another.

One of my notes from the conference reads:

“The way the battlespace moves you have seconds.”

Another:

“Need an integrated environment.”

And another:

“Quality v quantity problem.”

Taken together, they describe the real technological problem confronting modern intelligence.

America has become extraordinarily good at creating information.

Now it has to create meaning at approximately the same speed.

Ukraine's different answer

The discussions about Ukraine revealed the physical counterpart to this transformation.

Ukraine has been forced to innovate under conditions radically different from the procurement environment that produced much of the American military.

Expensive weapons still matter. Sophisticated platforms still matter. But so does something much simpler:

mass.

Cheap drones can be manufactured rapidly, modified through software, deployed in groups and replaced when destroyed.

Civilian production can become military production.

A person with a 3D printer can manufacture components. Commercial technology can be adapted for battlefield use. Small manufacturers can iterate designs according to what soldiers discover in combat.

The result is a form of civil-military fusion very different from the traditional image of a defense industrial base composed principally of giant contractors producing exquisite weapons.

Ukraine's experience raises an uncomfortable question for the United States:

Are we moving from bespoke warfare toward mass warfare?

The answer emerging from the summit was not entirely either/or.

America will continue to need what defense planners sometimes call “exquisite” capabilities: extraordinarily sophisticated aircraft, satellites, submarines, sensors and weapons whose performance cannot simply be replicated through volume.

But those systems may increasingly operate alongside enormous numbers of cheaper systems.

The future force could therefore resemble a barbell: exquisite capability at one end and mass-produced autonomous or semi-autonomous capability at the other.

The middle may be where the disruption occurs.

The economics of the drone war

Ukraine has also exposed a problem that is partly technological and partly economic.

What does it cost to destroy something?

If an inexpensive attack drone requires a vastly more expensive interceptor, the defender may successfully destroy the incoming weapon and still lose the economic contest.

The question therefore becomes not merely whether a system works but whether it can work at scale and at sustainable cost.

My notes repeatedly return to this problem:

“cost per kill.”

“low-cost mass.”

“counter-drone is more expensive than attack drones.”

“need to bend metal at scale.”

And, perhaps most importantly:

“barriers to speed — raw materials.”

Once warfare reaches enormous scale, supply chains become weapons systems themselves.

Motors matter. Electronics matter. Manufacturing capacity matters. Raw materials matter. Software updates matter. Trusted suppliers matter.

So do allies.

The future of military competition may therefore depend as much upon the ability to manufacture and replenish systems as upon the sophistication of the individual systems themselves.

The system becomes the weapon

Ukraine's Delta battlefield-management system provides another glimpse of this emerging architecture.

Rather than thinking about a drone, sensor or weapon independently, Delta attempts to bring information from different sources into a common operational picture.

That changes what military capability means.

The decisive advantage may no longer reside entirely in the performance of a particular platform.

It can reside in the connections between platforms.

A satellite detects something.

AI recognizes a pattern.

Another intelligence source adds context.

A drone confirms the target.

Software distributes the information.

A human makes—or supervises—a decision.

Another system acts.

The effectiveness of the entire chain depends upon how quickly and reliably those pieces communicate.

This is the same integration problem intelligence officials described throughout the conference, appearing in another form.

The battlefield is becoming a system of systems.

And the most powerful individual weapon may be less important than the architecture connecting thousands of relatively ordinary ones.

Where does the human remain?

That brings the argument back to the question hanging over almost every discussion of artificial intelligence:

Where does the human remain?

I wrote that exact question in my notebook during one session.

Immediately beneath it, I wrote:

“They didn't answer.”

But elsewhere at the conference, pieces of an answer appeared.

NSA officials emphasized human judgment, particularly for what one speaker characterized as no-fail missions. Humans remain necessary for legal accountability. They must understand the models and algorithms they supervise. They must monitor AI systems and recognize when the information being produced is unreliable.

Humans become, in one formulation from my notes, “overwatch.”

That is a subtle but profound change.

For generations, analysts searched through information themselves.

Increasingly, machines may search first.

The analyst then interrogates, validates, contextualizes and challenges what the machine has surfaced.

The scarce resource therefore becomes less the ability to collect information than the ability to exercise judgment over an enormous machine-mediated information environment.

That may explain another seemingly surprising theme at the summit: the repeated interest in younger workers.

Officials spoke about recruiting young analysts, IT professionals and technologists—not simply because the government needs more employees, but because younger “digital natives” may perceive technological possibilities differently.

One senior leader's attitude toward emerging technology was captured in my notes with an unusually revealing phrase:

“I'll try anything.”

For institutions famous for caution, that represents a remarkable cultural shift.

The lesson of 9/11 has changed

The attacks of September 11 appeared repeatedly in the discussions.

For the intelligence community, 9/11 remains the canonical warning about failing to connect information.

The phrase “connect the dots” still carries enormous institutional weight.

But the nature of that problem has changed.

In 2001, the lesson was that relevant information existed in different parts of government and was not sufficiently connected.

Twenty-five years later, the danger may be almost the inverse.

There can be too many dots.

Billions of observations, images, signals, transactions and sensor readings can theoretically be connected.

The challenge is knowing which connections matter.

That is why provenance, interoperability, data quality and AI repeatedly appeared beside one another at the summit.

Artificial intelligence can help discover relationships at a scale humans cannot.

But the faster machines generate those relationships, the more important another distinctly human question becomes:

Why should we believe them?

America's exquisite-machine problem

The United States built the world's most sophisticated military and intelligence architecture during an era in which technological superiority often meant building something an adversary could not build.

That assumption is becoming less reliable.

One sentence from my notebook may be among the most consequential observations of the conference:

“US no longer has monopoly on precision flight.”

Precision itself is proliferating.

Commercial components, cheap sensors, software, drones, satellite imagery and distributed manufacturing are allowing capabilities once confined to advanced militaries to spread downward—to smaller states, proxies and potentially non-state actors.

Iran's drone ecosystem illustrates one version of this phenomenon. Ukraine illustrates another.

The lesson is not that sophisticated systems have become obsolete.

It is that sophistication can now be challenged by simplicity multiplied thousands of times.

A $100 million capability facing one inexpensive drone has little difficulty.

Facing ten thousand inexpensive autonomous systems connected to sensors and software is a fundamentally different problem.

Intelligence at the speed of the machine

After two days of discussions about AI, satellites, drones, Arctic surveillance, data, targeting, analysts, commercial technology and warfare, the various subjects began to look less separate.

They are pieces of the same transformation.

Sensors create abundance.

AI filters abundance.

Networks connect the resulting intelligence.

Humans provide judgment.

Commercial industry supplies much of the underlying technology.

Distributed manufacturing creates physical mass.

Software coordinates it.

And militaries attempt to turn the entire architecture into action faster than an adversary can do the same.

The central competition may therefore be shifting away from possession of the single best platform toward something harder to photograph:

the ability to integrate an entire ecosystem.

That is why the intelligence community's problem and Ukraine's drone problem are ultimately related.

Both are trying to solve the same equation.

How do you take enormous numbers of relatively independent things—data points, sensors, satellites, algorithms, analysts, drones, manufacturers and allies—and make them behave like a coherent system?

Artificial intelligence will be indispensable to that effort.

But AI alone will not determine who succeeds.

The advantage will belong to the country capable of combining machine speed with human judgment, mass with sophistication, abundance with trust, and technological innovation with institutions capable of adapting quickly enough to use it.

The intelligence community is not merely entering the age of artificial intelligence.

It is entering the age of mass intelligence.

And that may require America to rethink not only how it analyzes the world, but how it builds power within it.

This article is AI-written from human-generated reporting, source material and editorial direction. The underlying observations and handwritten notes were produced by Ashley Greer while reporting from day one of the 2026 Intelligence & National Security Summit. Ashley used conversational AI prompts to interrogate the notes, identify recurring themes, test interpretations, develop the central argument and direct successive revisions. AI was used to assist with transcription and organization of the notes, thematic analysis, structural development and writing the article's prose. The original reporting, observations, questions, editorial direction and responsibility for publication remain with the author.