How AI and Drone Swarms Are Changing the Ukrainian Battlefield - Again

From one operator, one drone to one operator, multiple drones...to no operator. AI and swarm technologies, including 'terminal guidance' which gives the drone the ability to complete its mission without human guidance, are again changing the way war is fought by Ukraine against Russia.

The evolution and application of advanced technologies is actually making it harder to wage offensive war because the defensive success rates are so much higher. JL

Demian Shevko reports in New Voice of Ukraine:

It is increasingly hard to find weapons that don’t use it. AI is being integrated and applied very quickly — I wouldn’t hesitate to say everywhere. This affects everything: striking Russian troops at the front, how air-defense drones operate, and that automatic terminal guidance will increase ability to shoot down incoming Shaheds. The first swarm technologies are in use, where different effectors communicate with one another as one operator can control multiple drones. It’s already affecting the fight. If AI weren’t applied to drone technologies now, we wouldn’t have the mission success rates we see.

The head of Brave1’s defense cluster, Andriy Hrytseniuk, told Radio NV in an interview how AI-powered drones are evolving, why these technologies could make future wars impossible, which Ukrainian weapons attract Western interest, and which systems are currently being tested.

Radio NV: Many innovations are being developed with the help of BraveOne. Let’s talk about technological forecasts for the war — where it’s headed and what it might look like. How advanced are systems like terminal guidance, machine vision and automation now, Andriy? This affects everything: striking occupiers at the front, how air-defense drones operate, and claims that automatic terminal guidance will increase our ability to shoot down incoming Shaheds. Where are we in this process?

 

Andriy Hrytseniuk: If we speak about artificial intelligence broadly, it’s increasingly hard to find weapons that don’t use it. AI is being integrated and applied very quickly — I wouldn’t hesitate to say almost everywhere. Over these three years, the evolution in AI use has been astonishing.

If you look back a few steps, the early model was one pilot, one FPV drone, fully manual control. Engineers and operators then started asking how to increase effectiveness so a single pilot could engage targets more reliably — and how AI could assist that pilot. That led, for example, to terminal guidance. When a drone attacks a tank, there’s a technical concept called the radio horizon and control often gets lost. In footage of strikes you can see the picture degrade — the drone may approach from behind a hill or tree, or electronic warfare begins to jam the link and the operator can no longer control the drone effectively. That’s where AI comes in: technologies allow the drone, in a manner similar to an automobile on autopilot, to complete the flight and strike.

Those were the first steps in applying AI to drones.

 

discussions. In simple, non-classified terms: initially it was one pilot, one drone. Then AI started assisting in target engagement when the pilot lost control. The next step was one pilot controlling many drones — group missions. Often a single drone is insufficient to neutralize a high-value target, so several effectors are coordinated under one operator. Then there are technologies for automated decision-support and target marking so the pilot can see viable engagement options.

Beyond that we use the term “drone swarm”: many drones operating autonomously even without a human link. Technologically each level is far more complex, but dozens of Ukrainian teams and companies are working to develop these technologies so they’re available, user-friendly and scalable for the armed forces.

Read also: 

Russia copies Ukraine’s drone-interceptor advances – Syrskyi

So we’re at the stage where an operator can control multiple drones?

Yes — and the first swarm technologies are already in use, where different effectors communicate with one another. It’s not yet mass-scale, but it’s a matter of time.

When I speak with Europeans and Americans, there’s always skepticism. They ask: how can you allow AI to make lethal decisions — not a human? That’s a huge danger; it risks civilian casualties. We reply: we’re fighting people who hide behind civilians as human shields, and you warn us against automation because you might inadvertently hit your own relatives. How do you resolve that dilemma?

 

Any technology must include safeguards. I won’t comment on the exact safeguards or operational details, but systems are designed with protections to avoid strikes on inappropriate targets and to reduce the risk of drones hitting nonmilitary locations.

We observe that automation is advancing to early swarm capability. You say it’s not yet mass-produced. When can we expect this at scale, and what conditions are needed?

We typically announce technologies after they’re already in public use. Some publications and reports have shown that these technologies are being used, so we can speak about them, but not in deep detail — we avoid giving the enemy material they could copy.

Where is the enemy in this field? Do you analyze their use of AI? I ask because many participants say AI and network automation aren’t a “wonder weapon,” but they are the next phase of warfare and will dramatically affect the battlefield.

It’s already affecting the fight. If AI weren’t applied to drone technologies now, we wouldn’t have the mission success rates we see. This is a close contest: technologies are created both by us and, unfortunately, by the Russians. They have strong engineering teams and many scientists working on these problems. The question is who will be first, who will scale more broadly, and how Ukraine — with fewer resources than Russia — can develop asymmetric answers: unconventional approaches to more effectively strike the enemy.

We haven’t even reached 30, 50 or 80 percent of AI autonomy’s potential. Personally, I think AI is far more powerful than even nuclear weapons in some respects. There’s enormous potential; in theory, these technologies could make large-scale wars impossible in the future if an adversary were instantly neutralized. But that future requires many things to be accomplished before such capabilities are real and widely used on the battlefield.

A practical question that interests ordinary people who aren’t in weapons production, marketing or fundraising: why do we often discuss technologies publicly — missiles like Flamingo, Palyanytsia, Ruta, Peklo, Neptun, or drones like Drakons that received media attention last year? Why publicize that we have swarm capability? Does this help with funding or procurement?

Brave1 and the Ministry of Digital Transformation generally communicate only after a capability has been widely used, as I mentioned. A week ago there was a large event in Lviv showcasing Ukrainian technologies — nearly 300 companies presented, and on the Brave1 platform our cluster includes about two thousand companies. Some displayed products, others did not. So the idea that we “talk ahead” of development is not quite accurate. We discuss uses and publicized cases once it’s appropriate to do so.

Let’s discuss investment in Ukrainian armaments. Lately more Western defense companies have entered Ukraine. There’s interest, because Western air-defense systems were designed after WWII for intercepting high-end, relatively rare threats like ballistic missiles, not thousands of small drones per day. That’s why Polish interest in our air-defense drone technologies is logical. What are the global investment trends in Ukrainian weapons? Where are investors putting money and which development directions look most promising?

 

Interest goes beyond Poland. Ukraine has become a leader in innovative weaponry, and that trend is steady. EU countries, Britain, and the U.S. are actively studying our experience.

There are several trends. First: AI technologies — they generate huge excitement. Second: interceptors — a new class of weapon. Barely a year or two ago this didn’t exist; early 2024 saw experiments using interceptor concepts and deploying drones as air-defense effectors. This new class attracts global interest: for countries strengthening their own militaries and as an investment opportunity in Ukrainian companies. Foreign capital strengthens Ukraine’s economy, creates jobs and reduces the burden on Ukraine’s budget.

All innovative projects built with Brave1 grants were fully borne by the Ukrainian budget. New investments that are now entering Ukrainian companies reduce that burden, which is beneficial.

Some manufacturers worry that Ukraine could become an “arms China” — cheap labor producing inexpensive systems that Western firms will profit from by outsourcing production here. Is that a real risk?

Look at defense-sector wages: they’ve been rising monthly in Ukraine. I strongly disagree that this is cheap labor. While wages are lower than in Europe or the U.S., they’re high relative to other Ukrainian sectors and continue to rise. Even with higher wages, it’s still far more economical for foreign firms to develop new weapons here than at home.

If you consider the funds required to develop a new missile in the U.S., that could be hundreds of millions to billions. In Ukraine it’s orders of magnitude less for several reasons — wage is not the primary one. The key is that our engineering teams and companies can produce solutions much faster. Our specialists can do remarkable things. Ukrainian drones are becoming a benchmark for cutting-edge technologies worldwide.

You mentioned unique interceptor drone tech. What about developing our own surface-to-air missile systems? It’s crucial for us not to wait for foreign deliveries or for partners to reach their own technology cycles like SAMP/T or IRIS-T. Where are we in that process?

A few years ago missiles were the domain of state firms — slow and costly. Brave1 works with the private sector exclusively. Today over 40 Ukrainian companies are developing rocket armaments and many systems are already being tested. Waiting for foreign partners to produce new countermeasures against Shaheds would leave the economics of war stacked against us. If a Shahed costs, say, $100,000, the means to destroy it must cost far less.

Projects in Ukraine are developing surface-to-air missiles that are much cheaper than global analogs and effective against Shaheds. We must continually analyze the economics of war so our counters are cheaper than the adversary’s offensive means.

So our private manufacturers already have some anti-air solutions?

Yes.

When will we see them in action?

We already are seeing results. I expect that this year we’ll have verified successes in shooting down Shaheds using these systems, which will positively affect the battlefield situation.