Poland Will Buy Shield AI, if Shield AI Buys Into Poland

So far in the Defense Profile run, there’s been a lot of focus on individual companies as a lens on the defense industry and the global security situation.

But I want to focus on a more diverse range of topics to really understand how companies interact with nations, and how nations position themselves to manage their (often precarious) security situations.

I’ve been especially interested in the situation in the Baltic and Eastern Europe. These nations face existential threats, as they have through much of their histories.

What is encouraging and interesting, is the seriousness with which they approach their challenges, and the innovative ways they are adapting their economic and defense policy to stand tall.

So today’s issue takes the focus away from a pure focus on a company, and toward how those companies can and must work with their partner nations.

Poland is rearming faster than anyone else in NATO, and it's started charging more for admission to its market.

The price of selling weapons in Warsaw now goes beyond just a competitive bid. It's a factory, a technology transfer, and even a partnership.

And, as NATO's largest relative defense spender, Poland is in a strong position to demand more from those who want to do business there.

On Sept. 8, at the MSPO exhibition in Kielce, Shield AI signed a five-year letter of intent with Polska Grupa Zbrojeniowa (PGZ), the state-owned holding company that contains most of Poland's arms industry, and Wojskowe Zakłady Lotnicze nr 2 (WZL-2), its aircraft main aircraft works, to build the AI-piloted strike aircraft X-BAT in Poland.

It's still just a document. A letter of intent commits nobody to build anything. But it shows an approach Poland has spent three years developing: buy foreign, but require that a share of the work, the tooling, and the know-how stays behind in Polish industry.

Every defense purchase is meant to leave the country better prepared to make the next one itself. It's worked on tanks. Software might prove to be harder.

Poland spent 4.5% of GDP on defense in 2025, up from 3.8% the year before. In absolute terms that's about $44.3 billion, the alliance's sixth-largest budget. The government's 2026 plan pushes the figure toward 4.8%.

Nobody else is close. Lithuania sits at 4%, Latvia at 3.7%, Estonia at 3.4% — and those are the next most alarmed governments in Europe. The United States is at 3.2%, the United Kingdom at 2.4%, France at 2.1%.

Headline percentages can flatter some countries, because a national defense budget absorbs pensions, salaries, bases, and training as much as it actually buys anything physical.

Poland is different. It puts 54.4% of the budget into equipment, the highest share in NATO, which means the money shows up as orders for defense manufacturers.

That makes Warsaw perhaps the most valuable customer in Europe, and explains why vendors from Seoul to San Diego now pitch partnership and investment, as much as favorable pricing.

Urgency around defense isn't theoretical. On the night of Sept. 9, 2025, at least 19 Russian drones crossed into Polish airspace.

Dutch F-35s shot several down, Warsaw invoked Article 4, and NATO stood up Operation Eastern Sentry two days later.

It was the first time a NATO member had engaged Russian assets in its own airspace since the full-scale invasion of Ukraine in 2022.

European air forces are concentrated around a handful of main operating bases with few hardened shelters and thin missile defenses, a posture RUSI's Justin Bronk calls "highly vulnerable" to Russian long-range strike. Poland's sit closest of all.

Poland needs capability soon, which rules out building it all at home, and capability that survives a war, which rules out relying wholly on someone else's.

Recognition of this existential challenge, and of familiar ones faced in the past, have led to a clear resolve to bolster its defensive position not only by importing military capability, but through cultivating their own production capacity and technical talent.

Poland's last big tank order went to Hyundai Rotem: 180 more K2 Black Panthers , roughly a third of them will be a Polish variant assembled at Bumar-Łabędy's plant in Gliwice, with the first due in 2028. Polish firms will also help supply the navigation and optics.

Hyundai Rotem stays prime contractor. The design is Korean, the technology is transferred rather than owned, and nobody in Warsaw pretends otherwise.

This shows the shape of the rearmament. Poland isn't replacing foreign suppliers. In fact, it's buying from them at a greater scale than anyone in Europe. But it is insisting that a share of the building happens on Polish soil.

These will be the first tanks built in Poland since 2009. Nineteen years is long enough to lose a workforce, and rebuilding the line is as much about people who can weld a turret ring as the contract that pays for it.

Procurement from Polish and central European suppliers reached 30.4 billion złoty in 2025, close to four times the 2022 figure. Deputy State Assets Minister Konrad Golota put the policy to Reuters simply:

When it comes to cutting deals for tanks, the impact this will have on the defense industrial base, and the wider economy, is clear. It means steel, assembly lines, a trained workforce, and the ability to keep producing if supply lines are cut in a future conflict.

It also compounds. Each localized program makes Poland a more attractive base for the next one. This is the bigger goal. Warsaw isn't only trying to be Europe's biggest defense customer; it wants to be the continent's assembly floor.

X-BAT is an autonomous strike aircraft with a 1,000-nautical-mile combat radius, internal bays rated to 2,000-pound weapons, and 80 kilowatts of electrical power for radar, ISR and electronic warfare. A GE Aerospace F110 engine provides the power. Flight testing starts this year.

Its runway independence is critical. The ability to both launch and land vertically removes the dependency on airfields Russia has already mapped, addressing one of Poland's potential vulnerabilities in case of sudden attack.

As much as the X-BAT's sleek and eye-catching airframe, the product is Hivemind. This is Shield AI's autonomy software, the thing responsible for piloting it.

A conventional aircraft is a platform made useful only when a trained pilot is available, meaning its potential is unlocked by highly-skilled manpower alone.

An autonomous one arrives with judgment, experience, and instinct already installed, so the capability lives with whoever wrote it and how effectively they can update it.

The end state worth watching for is a Poland that is assembling these airframes domestically, but which it can't independently modify, since they are powered by American engines, and flown by American code.

That may prove effective, but it may also be a less dependable kind of military sovereignty than the K2PL tank program delivers.

The industrial logic is real. WZL-2, the Polish Air Force's main aircraft depot, is maintaining F-16s and C-130s, and it's already building maintenance capability for the F110 engine family used by X-BAT.

Zawieja's framing at Kielce was precise: X-BAT is "exactly the kind of advanced autonomous programme in which Polish industry should play a substantive role." Gazda pointed at WZL-2's "technical capabilities, infrastructure and highly qualified workforce."

Nor is anyone in Warsaw arguing for complete manufacturing independence at this point. As Defence Matters notes, locally assembled equipment "may still rely on imported engines, electronics or propellants."

Poland isn't trying to build everything so much as it's buying its way into the core production cycle of the weapons it may end up relying on.

All of this rests on a document that, up to now, binds nobody. The parties involved have still to settle workshare, facilities, technology-transfer terms, and industrial responsibilities before anything is signed.

The K2 playbook worked because the object was physical. You'll be able to walk into Gliwice in 2028 and count hulls. There's no direct equivalent proof software.

When the definitive agreement lands, does it give Poland the right to modify, certify, and sustain the software itself, or a license simply to bolt airframes together and buy updates from San Diego?

Tseng's line at Kielce was that "Poland is investing in advanced technology at a scale few countries in Europe match." That's true, and it's why the queue at MSPO was so long. The open question is what Poland does with the leverage.