Sometime in the early 2000s, the author of these lines participated as a Sherpa to Commissioner Chris Patten in the group called STAR 21, which was supposed to say something about the future of the European aerospace industry. Around the table sat the industry's biggest names from EADS, Thales, BAE and others, gathered to describe their own future.
Two conclusions from that work have proven to be sustainable, and they have become more accurate over the years. The first was that defense industries are increasingly less defense industries; their main tasks often lie in the civilian sphere or in the gray area where civilian and military cannot be distinguished. The second was that they are increasingly less uniquely European. They have roots in different parts of the world, not least in the United States, partly for selfish reasons: having American residency is a way to get around American export control restrictions, something that has hardly become less valuable during Trump's second term.
The thesis of this article is that what was an observation about Europe's large corporations in 2002 is today Sweden's structural conditions. There is hardly any uniquely Swedish defense industry left, and the exceptions that do exist are both fewer and smaller than the debate assumes.
What follows in this article is a compilation of detailed and important views that have been individually (on a Chatham House basis) presented by a number of prominent experts in the subject based on an AI-supported material collection. However, the final product has not been reviewed by all of them and the material is intended to be further developed within the project activities of KKrVA in collaboration with IVA.
How it happened: a chosen path, not an accident
This was not always the case. During the non-alignment era, the defense industry was part of the architecture of security policy: a broad domestic production capacity, from fighter aircraft to ammunition, would make the neutrality line credible, and the state controlled what was in practice an extension of the defense forces through orders, export controls, and informal ties. Defense decisions around the turn of the century turned the tide. When invasion defense was dismantled, orders shrank below the level that supported the breadth, and internationalization became official policy; anyone who reads the defense committees' reports from the late 1980s to the early 2000s can follow the turnaround almost year by year. The sales of Hägglund's (1997), Kockum's (1999), and Bofors' heavy weapons division (in 2000, to the American United Defense, which later became BAE) were thus not accidents in the work but consequences of a chosen line. It is the long-term effects of that line that Sweden is now living with.
What a defense company is, and no longer is
The weapons technology of the classic industrial era (tanks, artillery, armored ships) was primarily military with limited civilian spin-offs. The era of information technology reversed this. GPS was constructed militarily and civilized; the internet was built for military resilience but took on a civilian form; artificial intelligence and semiconductors are now funded and operated civilianly but are crucial militarily.
This has a consequence that is rarely drawn out. When the most important military technology is developed in civilian companies, the category of “defense company” becomes an administrative label rather than a description of where the capability is located. The question of who owns the defense industry is thus a subset of a larger question: who owns the technological base.
The Swedish map
Saab is the exception that is usually cited, and with some justification. At the end of 2025, Investor owned around 30 percent of the capital and 40 percent of the votes, and the Wallenberg foundations a further nine percent of the capital. Control is thus Swedish and stable, and there is much evidence that the Wallenberg sphere's grip on the defense sector is rapidly strengthening: through Investor, the foundations and SEB, positions are being built in several companies with defense relevance, in alliance with other wealthy Swedish ownership groups. At the same time, the Swedish layer is also more complex than it appears. The rest of Saab's ownership is spread internationally, with BlackRock and Vanguard among the larger institutional owners, and Saab's own growth is taking place to a significant extent through American establishment. In addition, there is a legal circumstance that is rarely noted: Saab's and Ericsson's operations in the US and Germany are registered as American and German companies respectively and are subject to the legislation of these countries to the same extent, with all that this may entail in a crisis.
If you go past Saab, the picture becomes different. The two second largest producers of defense equipment in Sweden, BAE Systems Hägglunds in Örnsköldsvik and BAE Systems Bofors in Karlskoga, are British-owned, which after Brexit also means owned outside the EU: the union's tools, from joint procurement to the review of foreign direct investments, only partially reach them. This is where the 90 combat vehicle and the Archer artillery system are manufactured, two of the systems most concretely associated with Swedish defense capabilities and with support for Ukraine. The ammunition manufacturing in Karlskoga also houses Nammo Sweden, which is owned in half by the Norwegian state and the Finnish Patria. According to a compilation by Svenska Fred, approximately 30 percent of the Swedish military equipment industry is currently foreign-owned.
Beneath the big three, Saab, BAE Systems Bofors and BAE Systems Hägglunds, a middle layer is growing that is easy to overlook because it is not yet visible in the major procurements. MW Group, with Stefan Persson's capital behind it, is building a fairly purposeful group; SVEKON, MilDef and W5 Solutions, among others, are growing significantly and in any case have ambitions to challenge the established ones in selected areas. What they have in common is that they are distinctly Swedish, in terms of ownership as well as operations, and that growth gives them something that Swedish defense companies in this size class have long lacked: access to capital. At the same time, the flow also goes the other way; Exensor in Lund, with a turnover of 400–500 million kronor, is part of a French group. In other words, the middle layer is not an unambiguously Swedish story, but it is where a new generation of Swedish-owned defense companies will be formed.
The submarines are the clearest counterexample, and the story deserves to be told in more detail than it usually is, because it contains two lessons rather than one. The government has explicitly identified two capabilities as essential national security interests: combat aircraft and underwater capabilities. Kockums was bought in 1999 by the German HDW, which was absorbed into ThyssenKrupp. Under German management, the shipyard was not allowed to offer its own products on the world market, and what Sweden was offered for the next generation of submarines was a submarine Sweden did not want. Without regained ownership, Kockums would most likely have been closed down as a newbuilding yard and the submarine weapon phased out. The 2014 buyback, for 340 million kronor, was in practice a hostile retake, a “hostile retake” that is quite unique in Swedish industrial history, and it was successful through concerted action between politics (in the midst of an ongoing change of government), the Armed Forces, FMV and Saab, ultimately the Wallenberg family. That is the first lesson: ownership determined which products the company was allowed to develop and offer, and thus whether the capability would survive at all. The second lesson became apparent later and is more inconvenient. The ability to effectively design and build submarines had already been quietly lost, without any of the actors, neither Saab, FMV or the Armed Forces, realizing it; the right expertise and infrastructure simply did not exist. That is a main reason why the A26 is taking twice as long to build, and consequently is twice as expensive, as it was intended. Today, the expertise is back, but it has cost ten years and several billions that no one counted on in 2014 and 2015. The repossession thus took place where the owner's calculation collided with a designated security interest; outside the two designated niches, there is no corresponding barrier, but the Kockums case shows what is at stake when the barrier is missing.
To this is added the large and often underestimated group: companies with a heavy civilian focus and significant defense relevance. The Volvo Group and Scania are among the most important, and they deserve a more prominent place in the picture than they usually get. Military logistics is a growing segment within both core areas, and it is through logistics that a modern war is financed. AB Volvo has an unusually broad portfolio under Swedish main owner control (Industrivärden is the largest): trucks, construction equipment, buses, marine systems through Penta and energy solutions, to a growing extent with automated and autonomous elements of direct military relevance. The ownership picture is also complex: Chinese Geely is the second largest voting shareholder in AB Volvo with just under 15 percent, and Volvo Cars is owned by the same Geely for around 79 percent. Interestingly, the groups are moving in different directions in China: AB Volvo has discontinued its own production there, while Scania, which is part of the German Traton, is once again investing in Chinese manufacturing. Axis Communications in Lund was bought by Japan's Canon in 2015. Ericsson went in the opposite direction and sold its defense electronics to Saab in 2006, thus becoming a civilian company with remaining defense relevance rather than a defense company with civilian operations.
Finally, below that lies a layer of small and medium-sized subcontractors where innovation is often the fastest and margins are the thinnest: individual companies with unique expertise, thin funding and practically no place in contingency planning. The chains are global in many layers, making them difficult to monitor even for the main supplier and accessible to cyberattacks.
Overall, it is difficult to point out more than a handful of significant activities that are both clearly military and clearly Swedish-controlled, although the growth of the middle class may eventually change that picture.
What ownership actually determines
It is worth being careful here, as the debate often jumps from ownership structure to conclusion without any intermediary. Ownership does not automatically determine where production takes place, where the expertise is located or who gets to deliver; BAE Systems manufactures in Örnsköldsvik with Swedish engineers, and it is Swedish export control legislation that determines where the products can go.
What ownership determines, however, are four things that are harder to see in everyday life. The first is investment decisions in a crisis: which facility is scaled up and which is closed down is decided by the group management, not by the municipality. The second is which legal regimes follow. An American ownership introduces ITAR regulation into the product chain, and a Chinese ownership introduces the reporting requirements that follow from China's legislation on civil-military mergers, which Sweden's national security strategy from 2024 points out as a threat to both security and competitiveness. The third is the vulnerability to intelligence gathering; the same strategy explicitly warns of systematic Chinese activity directed at the Swedish defense and space industry, research institutions and universities. The fourth, which the Kockums case most clearly illustrates, is the decision-making mandate over research and development: where development orders can be placed, what direction the company's own development work is given and which products the company develops, or is allowed to develop. This issue is easy to overlook in a boom, when orders carry everything. It becomes crucial the day order volumes slow down, for example if the Ukraine war, against all odds, ends in some form of settlement; that is when the group management's will to keep development capacity alive in Sweden is tested, and by extension the companies' ability to actually build the systems that the future requires.
However, the distinction deserves to be taken a step further, because not all dependencies are vulnerabilities, and not all vulnerabilities are strategic risks. A dependency only becomes a vulnerability when substitutes are lacking within the time a crisis lasts. A vulnerability only becomes a strategic risk when it can be controlled or eliminated by an actor whose interests may conflict with ours. By that yardstick, the Chinese near-monopoly on the strong magnets that are found in most qualified defense equipment is a strategic risk, while British ownership of Hägglunds is primarily a dependency. The ladder also provides a basis for sorting out politics: what is a strategic risk must be secured nationally or Nordicly, what is a vulnerability can often be managed at the European level, and pure dependencies are the normal state of the open economy, not a problem to be legislated away.
It is in this light that Geely's position in the Volvo sphere should be read. There is no public evidence that that position has been exploited in any problematic way, and that should be stated clearly. But the structure is there, and it is around companies that simultaneously supply vehicles and logistics to the armed forces.
What happens if there is war? On the ownership side, less than you might think, on other sides, more. Swedish disposition legislation gives the state control over production on Swedish territory regardless of who is on the share register. What cannot be commanded are the input goods at the far end of the chain, the engineers and the ownership group's willingness to invest in reconstruction afterwards. And seeking significantly greater control in advance than today is neither realistic nor unambiguously desirable; the price would be precisely the international access that makes the industry viable in peace.
Two of these factors deserve to be highlighted from the sidebar. One is the owners' will and feeling for the Swedish-based business. There is much evidence that physical Swedish owners are better placed to undertake long-term capability development, and to occasionally take calculated risks, than a group management in Paris, Berlin or London, not to mention the USA; the proximity creates a loyalty to the business that cannot be negotiated, despite the fact that Sweden is now a member of both NATO and the EU. The other is the competence base. The history of submarines shows that the engineering competence that can work closely with the user competence, in units and commissioning authorities, is vital for the survival of the industry, and that corresponding competence must also be maintained by the authorities in the role of commissioning. If it is not maintained through conscious decisions, in practice through a steady stream of development assignments, it will not help to invest money at a later stage; competence that has been dispersed takes a decade to build up again.
Two Swedish characteristics that make the issue more difficult
Two circumstances make these issues more demanding in Sweden than in many comparable countries, and they are not given sufficient space in the defense industry strategy that has been discussed anew between the state and industry players under the leadership of the Ministry of Defense.
The first is that the Swedish defense industry is entirely privately owned. Unlike France, Italy or Norway, the Swedish state owns no defense companies to govern; any ambition to govern must take a detour through orders, contracts and legislation.
The second is the administrative model. A constitution without a ministerial government means that the government cannot control the authorities that are supposed to monitor the situation and set requirements in detail; FMV, FOI and the supervisory authorities act independently within their instructions. The chain of control is thus doubly indirect: the state does not own the companies, and the government does not command the authorities. In countries where a minister can call a state-owned CEO, this happens more quickly. In Sweden, the same effect must be achieved through early coordination, clear mandates and awareness among all involved of where the real dependencies lie. That is a reason in itself to raise the issues now, while the instruments still have time to be built.
What's next for security policy?
The most common reaction is to demand resumption of ownership. This is difficult and perhaps not always appropriate, for two reasons. One is that Swedish capital markets do not accommodate the investment volume that a buyback policy would require without crowding out other things. The other is that international ownership also provides access: it is through the BAE ownership that Hägglunds accesses British procurements, and through the Nammo structure that Karlskoga sits in a Nordic ammunition system that is larger than what Sweden alone could carry.
However, a nuance is needed. In well-chosen niches, new establishment works on market terms: SWEBAL's trotyl production under construction in Nora, which is well advanced in the process, and the Wallenberg sphere's gunpowder initiative Verkan in Pjörneborg show that Swedish capital can support national investments where the bottleneck is clear and demand is secured. This is niche logic rather than buyback policy: it works where an individual critical input product lacks a sufficient European source, not as a general ownership strategy.
Against this background, five measures appear to be generally more effective than a broad buyback policy, even though the Kockums case shows that the ownership issue in the individual case can be absolutely decisive.
The first is to shift scrutiny from ownership to dependent. The question should not be who is in the share register, but what components, tools and software a production line cannot restart without. Parts of that picture already exist: FMV has good insight into the dependencies surrounding the materiel the agency itself procures, FOI and the companies see their respective sections. What is missing is the aggregation across authority and company boundaries, and above all the level beyond the first subcontractor, down to components and raw materials. Two reservations are crucial. The situational picture must not be built through processes that further delay the rearmament, and it has no value until someone has a mandate to act on it; seeing complexity is not the same as managing it.
The second is to take grey zone companies seriously in planning. If Axis cameras, Ericsson's network and Volvo and Scania's logistics systems are in practice parts of the total defence infrastructure, they should be treated as such in contingency planning, regardless of where the owner is registered. This in turn requires that the trust gap between authorities and business is addressed: understanding of each other's conditions is currently low on both sides, and without it, demands for security of supply can neither be reasonably set nor met willingly. The small subcontractors belong in the same planning; that is where both the fastest innovation and the thinnest margins are found.
The third is to use the regional layer consciously. The cooperation frameworks are now layered: NB8 with Poland as a permanent participant since 2024, Germany as a framework nation in the Baltics and driving force in the European Sky Shield, the British-led JEF in which Sweden has been a member since 2017, and the bilateral DCA agreement with the US since 2024. Industrially, this corresponds to a complementarity that has not yet been exploited: Norwegian ammunition and missile technology, Finnish ground combat industry, Swedish system integration. Expanding that logic, however, requires accepting that the companies in that system will not be national; Nammo is already an example of what such a system looks like when it works. The dependency ladder above indicates the division of labor: national and Nordic for the strategic risks, European for the vulnerabilities.
The fourth is to realize that export control, and in a broader sense that which secures priority delivery, is the real leverage. It was the insight from STAR 21 that companies establish themselves in the US to get around export control; the same mechanism works the other way, making Swedish and European control legislation one of the few instruments that actually bites transnational corporations. The fact that “ITAR-free” is today used as a sales argument by European manufacturers shows that the mechanism is already priced into the market.
The fifth is to nurture the competence base as what it is: a capability in itself. Development orders are not just procurement of equipment but the instrument that keeps engineering competence in the country and close to the users; the submarine episode shows what it costs to let it wither. This applies to industry, but to an equal extent to FMV and the Armed Forces in the role of purchaser, because a purchaser organization that no longer understands what it buys cannot control it either. If competence is not maintained with conscious decisions, money will not help at a later stage.
Is it a problem or an opportunity? The answer is probably both. The ownership issue is not settled by this; it remains important, and it will become more important again the day order volumes slow down and it is the will of the owners, rather than the order books, that keeps the development capacity alive. But the problem lies partly in a different place than where the debate is looking. The risk is not primarily that Swedish companies are owned abroad. The risk is that Sweden plans as if the category “Swedish defense industry” still described something delimited, and therefore neither asks the questions about dependencies, gray areas and legal regimes that the actual structure requires, nor the question of where the will and competence to build the next generation of systems should be located. That was, in essence, what the STAR 21 group was trying to say almost a quarter of a century ago.