Helsing vs Anduril: two defense AI companies compared

Two private companies with a similar pitch

Anduril is a private American company building autonomous defense systems, counter-UAS, and loitering munitions around a software-first approach. It sells an integrated command-and-control layer, called Lattice, rather than only airframes.

Helsing is a private German company that combines AI software for sensor fusion and targeting with mass-producible hardware, most visibly the HX-2 loitering munition line supplied to Ukraine.

Both companies pitch themselves as an alternative to the slow, cost-plus development cycle of a traditional defense prime, and both remain privately held with no listed common stock. The comparison that matters for a researcher is not which company is bigger today. It is which government market each one is built to serve, and whether its stated business model is software licensing, hardware production, or both at once.

Company Snapshot

Anduril and Helsing both describe themselves as vertically integrated: a proprietary software layer paired with hardware the company builds itself, rather than a pure software vendor selling into someone else's airframe. The difference is geography and customer base. Anduril's disclosed government work runs through the United States Army, the United States Marine Corps, and the Royal Australian Navy. Helsing's disclosed customer base centers on Germany, the United Kingdom, and Ukraine, with a reported effort to enter Japan through a brokered partnership.

Why the two are compared so often

Anduril and Helsing show up together because they are the clearest test case for whether a software-first, privately funded defense company can out-compete a traditional prime on speed and unit economics. Both argue that a mass-producible, software-updatable system beats a small run of exquisite, hand-built platforms. The pitch is the same. The customer base is not.

Anduril's government work is documented across three branches of the US and Australian militaries in named contract announcements, which is the strongest kind of evidence a private company can offer absent a public filing. The enterprise IT and software contract awarded in March 2026 carries a ceiling of up to $20 billion and integrates the Lattice platform for counter-UAS work tied to Joint Interagency Task Force 401, though a ceiling is the maximum the government may order, not a guarantee it will.

Helsing's public record leans more heavily on wartime deliveries and press-reported funding rounds than on named, dollar-denominated government contracts of the scale Anduril has disclosed, which reflects Europe's more fragmented procurement structure as much as it reflects company scale.

Software licensing versus mass production

The two companies also split on what they are actually trying to sell at scale. Helsing's stated pitch to European governments is software-defined, lower-cost, rapidly scalable systems instead of exquisite low-volume platforms, and the open research question is how much of its business is software licensing, through Altra and Centaur, versus hardware sales through the HX-2 line, because the two carry very different margin and scaling characteristics.

Anduril's public contracts suggest more revenue currently comes from integrating its Lattice software into government-owned or government-funded hardware programs, alongside its own aircraft and maritime systems, rather than a single dominant product line.

Centaur, Helsing's AI-piloted fighter software, has been tested aboard a Saab Gripen, which points toward a strategy of selling autonomy software that can fly inside an existing allied airframe rather than requiring the customer to buy a new Helsing-built aircraft. That mirrors Anduril's own bet that Lattice can run across more than one partner platform. Neither claim is fully proven in production yet, and reported hardware and delivery volumes on both sides should be checked against a government customer's own confirmation rather than a company statement alone, since attritable-drone production claims are difficult to verify independently.

What would change the comparison

A confirmed initial public offering on either side would be the single biggest change to this comparison, because it would convert a set of press-reported funding rounds and government contract ceilings into audited, filed financial statements. Until then, a headline valuation reported in trade press is a fundraising data point tied to investor sentiment, not an independently audited number, and it should not be read as a revenue or backlog figure for either company.

A second thing that would change the read is confirmation of Helsing's reported expansion into new markets outside Germany, the UK, and Ukraine. A confirmed Japan deal, brokered through Rakuten or otherwise, would extend Helsing's customer base into the Indo-Pacific and narrow the geographic gap with Anduril, which already has a confirmed Australian customer through the Ghost Shark program. The reverse also holds: if Anduril's enterprise IT ceiling converts into only a small fraction of obligated task orders, that would weaken the case that Lattice has become the durable, cross-program software layer its backers argue it is.

How to Use This Comparison

Neither company offers a public-market route today, so this is a competitive-landscape comparison rather than an investment comparison. Investors looking for listed exposure to the same themes, software-defined autonomy, counter-UAS, and loitering munitions, should look at the publicly traded suppliers and primes that compete in the same mission areas, rather than waiting on either private company to list. Track disclosed government contract announcements directly from the awarding branch or department, not secondhand trade coverage, before treating either company's growth narrative as confirmed.

For listed exposure to the same competitive set, the counter-UAS research page names the public suppliers competing against both Anduril and Helsing in detection and mitigation layers, and the collaborative combat aircraft page covers the qualified autonomy vendors, including public defense primes, working the same software-defined-aircraft problem Centaur and Lattice are both built to solve.

A separate caution applies to how these two companies get framed in coverage. Helsing is frequently described as Europe's answer to Anduril, and Anduril is frequently described as the American company European governments are trying not to become dependent on. Both framings are shorthand for a real sovereign-capability argument, not a technical or financial claim about either company. Treat a claim about market position as a claim to verify against a named contract or a company's own release, the same evidence standard this comparison applies throughout, rather than as settled fact because a headline states it confidently.