E 767 - Unicorn Atlas #1: Helsing - Europe's $18 Billion Defence AI Bet

On the 13th of July, 2026, a defense startup that did not exist five years ago, closed the largest venture round in European defense tech history, $1.8 billion, $18, billion post-money valuation, larger in a single raise than most European defense primes make in a year of profit. This company sells kamikaze drones to Ukraine and now to Berlin. It says it's predominantly European owned.

Its two biggest new investors are American. This is the first Unicorn Atlas entry. This is Helsing. Hello and welcome everybody.

This is episode 767 of StartupRate.io recorded by me, Jörn Manninger, aka Joe from Startup Radio, from Frankfurter Main, Germany. And this is the first entry in a new series here on the channel. We call it the Unicorn Atlas.

It will mostly be blog posts, but occasionally we also use the opportunity to introduce you to audio and video to a new unicorn, to an existing unicorn, or to one that, maybe even failed. We'll see. Every entry in the Unicorn Atlas takes one European unicorn and does what most coverage of those companies does not do. It asks who owns them, what they actually make, what is the product, why they are so valued, whether the numbers behind the headlines hold up, and what they mean for the European operator, investor, and policymaker who has to make decisions about the world these companies, are building.

Unicorn Atlas number one, Helsing, Europe's most valuable pure play defense tech company, also on paper Germany's most valuable unicorn startup, the one, European defense industry conversation now has to root through. And as of the 13th of July, an 18 billion dollar business on paper. Over the next roughly half hour, I'm going to walk you through what the Series E actually is, who put, in the money, what housing actually makes, where the money is going, what the Bundeswehr framework contract really means, why Ukraine matters more than most Western coverage admit, and most importantly, where the valuation stops being a bet on capability and starts being a bet on wartime.

Let's go! On Monday the 13th of July 2026 Helsing announced that it had closed a Series E round of 1.8, billion US dollars at an 18 billion dollar post-money valuation. Defense News confirmed the same number.

The German DPA, Deutsche Presse Agentur, German Press Agency were confirmed the same number. There is no ambiguity on the headline. Here is what makes it interesting. CRSD was one year earlier, in June 2025, 600 million euros at 12 billion euro post money.

Prima Materia, Daniel Elk's investment vehicle, the Spotify founder, led it. 12 months later, the valuation has gone from 12 billion euros to 18 billion dollars. Call it roughly 16 and a half billion euros. That is 15% up on a company that was already the most valuable defense startup in Europe.

For context, 15% up in 12 months. At this scale, on a company that publishes no revenue figures, is not the shape of a growth stage round. It is the shape of a scarcity trade. There is exactly one European pure-play defense AI unicorn of Helsing size and product breadth, and every institution that decided in 2024 or 2025 that European defense was a real thematic allocation, now has one target to buy in.

That is how you get 18 billion. Now here's the first thing that most coverage got wrong and here is where I want to start being useful. In May 2026 several outlets reported that Helsing was closing a 1.2 billion dollar round led by Dragonair and Lightspeed.

Some of you read those reports you may have wondered whether the July race was the second round on top of that it was not the May reporting and July close are the same event the May headline was the anchor tranche of what became series E the July announcement is the same round after final syndication expanded the ticket to $1.8 billion. There is no separate $1.2 billion round.

There is one Series E closed at 13th of July at $1.8 billion US dollars. Anyone building a housing timeline should merge those two lines into one. On to who puts in the money, because this is where the interesting European sovereignty question sits.

Lead Investor, Dragoneer Investment, American, based in San Francisco, Co-led Lightspeed Ventures, American, based in Mandelau Park. Other participants in the syndicate, per the official press release and Reuters coverage, Goldman Sachs Growth Equity, JPMorgan Chase, CPP Investment, that is Canadian National Pension Plan, General Catalyst, Iconic Plural Stepstone, Disruptive Ventures. Existing anchors Prima Materia, Excel, Green Oaks remain shareholders. Whether they participated in the Series E as new capital was not publicly disclosed.

Helsing itself uses the language predominantly European-owned, which is a careful phrase. It does not say the round was European-led. It says the cap table, the accumulated ownership from Series A onwards remains majority European. Both statements can be true.

Because when you do the arithmetic, $1.8 billion, going in at $18 billion post money implies that the newly issued shares represent roughly about 10% of the company. Existing shareholders are diluted by roughly 10%. So if before CSE the cap table was, say.

95% European after Series E, it's roughly 86% European. The company can still fairly claim predominantly European owned, while the marginal new capital, the incremental power, sits in San Francisco, New York and Toronto. I want to be careful here. One number you may have seen that existing chain holders retain 80-85% of the company post Series E, does not hold up to the arithmetic.

If you inject 1.8 billion at an 18 billion post money, the incoming investors get 10%, not 15 to 20. Existing shareholders keep roughly 90% of the equity minus whatever option pool refresh happened, which is not disclosed. The public math points to roughly 10% delusion.

Anyone quoting a bigger number is either double counting a secondary or extrapolating from CRSD. It is worth flagging because this misreading is now traveling in some coverage. The important political point is Europe just built its flagship defense tech unicorn on American and Canadian institutional money. That is not automatically bad.

Pension funds money is patience, well-priced and structural. But it does mean when European ministers describe Helsing as an emblem of European strategic autonomy, the equity behind the emblem was underwritten by non-European institutions taking a bet on the European defense spending surge. That is the kind of nuance the Unicorn Atlas is here to hold. Not to condemn it, to name it.

Onto what Helsing actually is. Three founders, Munich Orangen, founded 2021. Converted from a GmbH, a German LTD, to an SE, the European, company legal form in 2025, which is itself a signal about a company preparing for cross-border European growth. Thorsten Reil, co-CEO, his previous company was Natural Motion, the games and animation tech startup behind CSR, Racing, sold to Cigna, in 2014 for around 527 million US dollars.

That is why Daniel Ek, who has known Rael since the games area, was the first serious institutional check into Helsing in 2021. That is not a defense industry pedigree, that is consumer tech pedigree that pivoted into defense and is not worth naming because it changes how you should read Helsing culturally. Helsing behaves more like a US style hyperscale defense startup than like a European prime. Gunbert Scherf, Co-CEO, ex-Bundeswehr, ex-Defense Intelligence, worked on defense procurement and AI-adjacent programs inside the German security establishment.

He is the interior of the Bundeswehr Rolodex. He is why the Bundeswehr framework contracts happen as fast as it did. He is the German establishment site of the founder table. Dr.

Niklas Köhler, President and Chief product officer, machine learning engineer, founded an AI startup called Hellsicht that was folded into Hellsing. He is now the technical face. Now to the product taxonomy. This is where the 18 billion US dollar defense tech question actually gets real.

Hellsing does five things. 1. HX2, a loitering munition. Call it a kamikaze drone if that is easier.

Roughly 100 km range, AI-powered target recognition, designed to be resistant to electronic warfare. In serial production at the Munich Resilience Factory at over 1,000 units per month. That is the product paying the bills. 2.

Ultra, the mission command software layer. This is the actual AI product. Fuser sensor data generates target assignments, coordinates drone swarms and critically can operate in GPS denied environments, for example where Russian jamming has knocked out satellite navigation. If you strip housing to one durable mode it is ultra everything else is a container for ultra three, CA-1 Europa, an unnamed AI fighter jet concept, prototype unveiled in 2025, first flight planned around 2027.

This is the moonshot. This is the storyline that justifies the $18 billion valuation. 4. SG-1 Fathom, an autonomous underwater glider, long endurance underwater surveillance prototype tested in 2025, zero production in Plymouth, UK.

This is the maritime bed and it is the one European naives are quietly interesting in. 5. A range of adjacent programs. The Centaura Unnamed Escort Concept, the KIRA Electronic Warfare Anti-Submarine Program, where public disclosure is still thin.

Behind those five, there's a corporate manufacturing stack, Grob Aircraft for Composite Airframes, Blue Ocean Marine Tech for underwater vehicles, which we will get into. And there is a partner stack OHB and Kongspark and Hensoldt in space, Jurenko in warheads, Eurenko, sorry for the pronunciation, Instalaza in tactical air to ground munition. That is the company. Now let's follow the money.

Helsing's own press release says the Series E Capital is going into three things. One, integration of new AI platforms into partner nation defense systems. Two, expansion of the production footprint, the so-called resilience factor. Three, the CA1 Europa Unmanned Fighter Program.

Let's translate that into observable facts. The first resilient factory sits in the Munich area and produces over 1000 HX2 units per month. The company has publicly said its ambition is to reach millions of units per year across the network. The second facility is in Plymouth in the United Kingdom tied to a 350 million pound joint investment package announced with the UK government focused on both the HX2, and the Fathom underwater collider.

The third, and this is one that changes the geopolitics, is in Princeton versus Virginia in the United States. Housing is opening a resilience factory on U.S. soil that was announced as part of the Series E press cycle.

This is actually not a small decision. That is, Helsing saying it wants to sell into U.S. Defense market, competing directly with companies like Undruly and Shield AI on their home turf with local production for reasons that are much more regulatory than they are logistical.

US defense procurement heavily favors domestic manufacture. A European unicorn opening an American factory to sell to the Pentagon is a story you would have called improbable in 2021. On ca1 europa first flight is targeted around 2027 that is roughly 18 months from now if it flies the 18 billion valuation looks conservative if it slips two or three years which is the base rate for military aviation programs at any company including the primes the run may math may change fast, This is the money, now the contract that everyone talks about.

In February 2026, the German Bundestag approved defense procurement vehicles, the so-called 25 million Euro a piece Vorlage. Authorized framework contracts for two loitering ammunition supplies to the Bundeswehr Helsing with the HX2 and Stark, the other German defense tech startup, with a virtuous drone. The Helsing framework up to 4,300 HX2 units, framework ceiling roughly 1.4 billion euros.

The stock framework, 2,200 virtuous units, framework ceiling, roughly 2.86 billion euros. Those are the numbers you've seen quoted. Here is what most coverage does not say clearly.

A framework contract in German defense procurement is a ceiling, not a commitment. It is the maximum the German military, the Bundeswehr, may spend against the supplier over the contract's lifetime if it chooses to. The actual money that flows is triggered by call-off orders, or abrufe, each of which has to be individually approved and budgeted. The first Bundeswehr call-off from Helsing framework was reported at approximately 270 million, with an M, euros, not 1.

46 billion. 270 million. The rest is optionality. It is the option to spend up to that ceiling, an option that will get exercise or not depending on how the Zeit & Wende defense spending.

Posture holds up under future coalitions, future budgets and future assessment of the Russian threat trajectory. I want to be direct about why this matters. If you read the financial press and see Bundeswehr signs 1.46 billion contract with Helsing, and you build a revenue model on that number, you're modeling optionality as revenue.

When Helsing values itself at 18 billion, part of that valuation is a bet, that this optionality gets exercised at close to 100%. This, that is possible, but it's not guaranteed. And it's the single most important asymmetry an investor evaluating this company should hold in mind. The Ukraine side of Helsing's revenue equitation is more specific.

That is where we go next. Since 2022, Helsing has been supplying Ukraine, not as a side project, as the operational reality on which the entire product credibility of the HX2 is built. By early 2026, Helsing has publicly. Claimed several hundred HX2 units delivered per month to Ukraine forces.

Defense Network, a German language defense industry outlet, reported that in the winter of 2025 to 2026 alone, over 1,000 HX2 units were delivered to Ukraine. The Ukraine armed forces have used them in what Helping calls logistic lockdown operations, hitting Russian rear area supply. Ukrainian commanders are also embedded in Helsing's product development loop. The Ajax 2 is iterated in something close to real time based on field feedback.

For a defense tech company, this is genuinely unusual. Most defense primes iterate on multi-year procurement cycles. Helsing is iterating in weekly loops in a live war with a customer who tells them exactly what does and does not work. That is the mode that money cannot buy.

And this is the single most important piece of Helsing's product credibility. But, and this is where the report to financial caution I want to bake into every Unicorn Atlas entry has to enter. The operational picture is not uniformly positive. In January 2026, Bloomberg reported that early German Army Ajax-2 test flights had failed.

Systems were failing to arm properly, navigation modules were reported missing, and initial reorder cycles were paused pending resolutions. Helsing publicly disputed these reports and cited near 100% hit rates on its own operational data. I do not have a way sitting in Frankfurt with public sources to arbitrate between Bloomberg and Helsing. What I can tell you is this, both things can be true.

Loitering munitions of this class are notoriously difficult to operationalize outside their proving ground. The HX2 was hardened by Ukraine on Ukraine terrain against Russian jamming patterns. When you take that product and drop it into German army field exercises in Central European terrain, some fraction of it does not port cleanly. This is not a scandal.

That is the normal operational reality of taking a battle-hardened product and standardizing it across a NATO customer base. The question the unicorn asked us has to hold and the question the operator investor and policymaker listening to this episode has to hold is whether helsing's engineering, can converge those two performance regimes fast enough that the bundeswehr confidence stays intact through the framework's contract call-off cycle right now the answer is probably not certainly here is where helsing gets more interesting than a single product story.

Over the last 18 months Helsing has been building through acquisitions, joint ventures and MOUs something that looks like a sovereign European defense tech supplier stack. Let me talk you through it. Grob Aircraft, acquired June 2025, located in Tussenhausen, Bavaria, roughly 275 employees, historically a maker of training aircraft and composite airframes. Helsing bought them for one reason, composite manufacturing capability for airframes.

This is the industrial base under the CA1 Europa. Plu Ocean Marine Tech acquired May 2026 roughly 120 employees, UAV, Autonomous Underwater Vehicles Engineering. This is in the industrial base under the SG-1 Fathom and the broader maritime program. KERC joint venture announced May 2026 with OBH in Germany, Koenigsberg in Norway, and Hänsoldt in Germany.

Purpose Tactical ISR Intelligence Surveillance Reconnaissance Satellite Constellation. This is Helsing telling European governments you do not need to depend on Maxar or Starlink for battlefield satellite imagery. We a European consortium will build the sovereign alternative. Eurenco.

Eurenco French warhead manufacturer MOU You signed June 2026 purpose, European sovereign explosive payloads for HX2. This matters because it lets Helsing tell customers their HX2 warheads are not subject to. USITAR, the American export control regimes that has reportedly become a political constraint for European defense exports. INSTALAZA, Spanish technical munitions supplier ongoing corporation on air to ground weapons.

And sorry for butchering our company names. If you stack those partnerships up, you can see the shape of the bed. Helping it not just building products. It is building a European defense tech ecosystem around itself in which it sits at the center.

That is exactly the new prime playbook. Andruly has been running in the US since 2020. And that is where the 18 billion valuation stops being outrageous. Let me put Helsing in its competitive frame because this is the question every serious observer of European defense tech is asking.

Andruly, in the US, most recent valuation on public records was around 14 billion dollars in 2023, likely higher now. Shield AI, US, valued in the low billions. Palantir, US publicly traded, market cap north of 30 billion. But Palantir is more of a data analytic prime than a hardware startup.

So the comparison is too loose. Rebellion Defense in the UK, smaller and quieter. On the European side, Stark, valued in the low single digit billions of euros, mainly on the Virtus program. ICIYE, the Finnish synthetic aperture satellite company above 2 billion euros.

Harmattan ai in france earlier stage, helsing at 18 billion dollars is the largest of these and the question is is that a bubble, here's my honest read in pure defense of tech unit economics cost per drone margins per system revenue predictability of framework contracts no at 18 billion helsing is not obviously trading on today's fundamentals helsing does not publish revenue if helsing's revenue in, 2026 were hypothetically in the 400 million euros to 700 million euro range range i'm inferring from a german call-off in the ukraine delivery flow and the emerging us and uk contracts then 18 billion is somewhere between 25 and 40 times revenue.

Well, you know, as a multiplier for the hardware heavy defense business, this is expensive. But the new prime thesis is not a revenue multiple bet. It's a bet that European defense spending search site and vendor. We are in Europe.

The 2% GDP NATO floor becoming a 3% floor in some countries creates a 15 to 20, year demand curve that Helsing is best positioned to capture in software-defined domains if that thesis is right 18 billion is a floor not a ceiling if this is wrong if European defense spending peaks in 2028 and rolls back 18 billion is a overshot, I do not know which of these is right neither honestly do the American investors who led this round what they are pricing is optionality on the first scenario.

This is what you see. At series e a scarcity market you pay for the option act nine ownership governance and a predominantly european question, let me come back to the ownership question because this it is, the piece of the story that european coverage has been softest on the two lead investors dragoneer and lightspeed are american goldman sachs growth equity jp morgan chase cpp american american Canadian. These are institutions with fiduciary duties to delimit partners and their pension beneficiaries.

It's not the European strategic autonomy their duty is to return. That does not mean these investors are hostile to European interest. It simply means government's interests are their own. And in a company that is now the flagship European Defence Tech Unicorn, That is a governing question that European policy makers and the German government should be, and I suspect are, asking privately.

Helsing says the cap table remains majority European. That is likely true. What is also true is that when you compound roughly 10% dilution rounds like this one over the next two or three years, because Helsing will raise again if the CA1 Europa program is capital intense, the European majority steadily thins. There is a version of the story in which Helsing goes public in 2028 or 2029 and is listed on a US exchange because that is where the liquidity is.

There is a version in which Helsing acquires or is acquired by a US prime. There is a version in which the SE structure gets structured for tax and regulatory reasons that shift decisions authority. None of these outcomes is imminent. All of them are on the range of scenarios that a European policy maker looking at this round should be modeling.

And that is the third thing the UNICORN Atlas is here to name. Not to editorialize, to make sure the question is not skipped. Let us give you our little verdict. Let me try to compare everything I have just said into a verdict, because the Unicorn Atlas has to close with one.

Helsing is a real company. It has three products operating in a real war. It has a real Bundeswehr framework contract in which the first real call-off has already been signed. It has a real acquisition and partnership stack that looks like the beginning of a European defense tech ecosystem.

It has three founders who between them combine consumer tech scaling experience, German defense establishment credibility and machine learning depth. That is not a fake company. That is not a wartime marketing exercise. That is a real business.

Helsing is also a company prized on optionality. Optionality on the Bundeswehr framework being exercised close to its ceiling. Optionality on CA1 Europa flying Optionality on Ukraine remaining a proving ground long enough for the HX2 to accumulate the kill data that justifies the software mode, Optionality on the European defense spending surge lasting 15 years Optionality on the American-led syndicate's willingness to support future rounds, If most of those options are exercised favorably 18 billion looks conservative.

If two or three of them break unfavorably, 18 billion looks like a wartime peak. My verdict is this, housing is a real thing. It is also the most exposed to political and operational risk of any European unicorn currently on the market. That is not a contradiction.

That is the specific character of a defense tech unicorn born in the middle of a European rearmament cycle. The question the operator, the investor and the policymaker listening to this episode should be asking is not, is housing overvalued? And it is, am I comfortable with which of these optionalities I'm underwriting when I engage with this company? For the operator, Ajax 2 is a real product.

Ultra is the mode. Take those seriously. Take the CA1 Europa story as future value, not as present capability. For the investor, watch the second and third Bundeswehr call-offs against the framework.

work. Watch the West Virginia factory first US contract. Watch the CA1 timeline slippage. Those three data points will tell you whether the 18 billion valuation is holding.

For the policymakers, the predominantly European-owned phrase is doing a lot of work. Ask harder questions about future dilution, exit venues, and governance rights of the American co-leads. This is the flagship of a strategic sector, its Ownership Org is a matter of European industrial policy. That is the first entry in the Unicorn Atlas.

If you found this useful, please rate and review Startup Radio wherever you're watching or listening. The companion blog post with the useful data tables, the funding timeline, the founder dossiers, the resilience factory, geography, and the full source list is on our blog startuprate.io forward slash blog. There's also a linked reading list in our recent Europe coverage.

The June 2026 startup news episode on the defense capital supercycle, our earlier piece on European structural recovery, our German VC episode, episode 762 on why stability is not strength, and episode 764 on the German AI bottleneck. Next entry in the unicorn atlas will be European unicorn with headlines valuation and the underlying business need to be pulled apart carefully stripped, subscribe on YouTube Apple podcast Spotify and wherever you find your audio and I will see you in the next one this has been your manninger see you soon.