A new prime, the same lock-in?
Where uncrewed vessels operate at scale, the manufacturer currently runs support. If that is scaled to a complex warship, then is support just locked into a new prime? The support choice is not between old or new prime, but between what is built on the supplier's architecture and what is built on the customer's data.
The last article looked at how the manufacturer maintains the small autonomous vessels, operating it from the shore and holding the data that says how it is performing. How does this scale to larger autonomous ships, and on whose terms?
Locking support into a single supplier is one approach, and has arguably been the 'norm' at a system or equipment level through UK defence procurement for some time. This article seeks to promote debate about whether the move to a Hybrid Navy with uncrewed assets will see sustainment locked in more deeply, or provide an opportunity for a more open approach.
The Defence Industrial Strategy of September 2025 is 'designed to prioritise UK-based businesses to deliver the Defence Dividend' and wants 'assured design, development, production, maintenance and support capabilities in the UK'.[1] It commits the Ministry to spending £2.5 billion more with small and medium-sized firms by May 2028, a 50 per cent increase.[1] The Defence Investment Plan of June 2026 adds 'buy British by default', an offsets regime and more than £5 billion for autonomous systems over four years, and gives the new Uncrewed Systems Centre the job of setting 'common standards, open architectures, data frameworks, and assurance models'.[2] The Defence Industrial Strategy appears to be silent on intellectual property, open architecture, through-life support or the right to repair,[1] meaning the strategy that seeks a broad UK supply chain appears silent on the one thing that decides who may support a system once it is in service: who owns the data?
The landscape
Support for the Royal Navy's ships is already an enterprise endeavour. The Future Maritime Support Programme (2021) was ten contracts across three naval bases, and has been extended from March 2026 to October 2028 while its successor, the Naval Support Integrated Global Network, is designed as a framework of contracts.[3] In addition to the platform-level support arrangements, there is a vast array of equipment and system contracts with equipment manufacturers, third-party support providers, and spares provisioning aggregators.
In 2021, the NAO found that the Ministry 'relies on a limited specialist supplier base', that 14 of the 20 projects it examined were 'wholly or partly procured non-competitively', and that about 44 per cent of procurement spend went to ten providers.[4] Support is not a minor part of this, with the ten years of the 2021 Equipment Plan showing that the MOD expected to spend £115.7 billion on support against the £114.5 billion allocated for buying the equipment in the first place.[5] Whoever holds the support holds half the money, and although the NAO's finding is about procurement, the supplier base it describes is the one that holds most of the support contracts too, many of them on long single-source arrangements. In other words, ‘lock-in’, is not an emerging risk from new acquisition programmes, but a feature of the landscape today.
Supply chains published, support models not
With the narrative that surrounds the neo-primes and defence reform, a legacy defence prime might feel uneasy about the threat to the market share they already hold. For mid-tier companies and SMEs the threat will likely be heightened, as if they don't feature as part of the supply chain today, what chance is there with a new set of primes emerging? It is worth exploring this further, as some of the neo-primes are very good at publishing their supply chain.
Anduril UK has a partnership with GKN Aerospace and a list of named British SMEs for its Army programme, and in April 2026 it partnered with Kraken, a UK builder of small uncrewed boats, for manufacture in the United States under licence (although the investment, the exclusivity and the integration split were not disclosed).[6] Anduril's Lattice software is integrated through a developer programme open to 'qualified developers' rather than through a published standard[7] The partnerships are a demonstration of a neo-prime embracing collaboration with a broader ecosystem; the interface itself is not yet published in the same way.
Similarly, Saronic, which is building a USD$3 billion shipyard of its own in Texas, has gone to Vigor (a long-established American ship repairer) for the maintenance, repair and overhaul of its vessels.[8] Even the neo-prime that builds its own hulls buys its depth support from an incumbent, offering some potential for a more vibrant market for at least Float and Move.
Where the terms are visible, support appears to follow the manufacturer as a conscious choice. Australia's AUD$1.7 billion Ghost Shark contract with Anduril is for 'the delivery, maintenance and continued development' of the fleet over five years, with more than 40 Australian companies in the supply chain.[9] What that contract does not remove, however, is the customer's challenge. When asked at the press conference that followed how a stretched navy would handle the new capability, the Chief of Navy answered that 'it doesn't require a crewing capability, it's just the sustainment challenge for us'.[10] Maintenance is written into the manufacturer's contract, and the Navy still sees sustainment as its own challenge, where the bundled contract moves the work, not the accountability for a fleet that has to be available. Australia has chosen to place the work with a supplier for a fleet of autonomous submarines, whilst writing its own industry's participation into the same contract. That shows a national supply chain and manufacturer-led support are compatible. What Australia did not attempt was a competed support market, and that is the half of the UK's ambition its strategy is silent on.
Buying the service rather than the ship
The most extreme version of the OEM model is one where the Navy does not own the vessel at all. The US Fifth Fleet's Task Force 59 ran Contractor-Owned, Contractor-Operated boats under Naval Oversight (COCONO) and the then Secretary of the Navy said there was 'no reason why contractors can't actually own this, sort of like a software-as-a-service model'. However, he also noted that it suited 'ISR or other non-lethal activities', and for weapons platforms that 'the US government will own those and control those unquestionably'.[11] The UK draws a similar line, with the MOD's stated approach to AI-enabled capability being that 'there must be context-appropriate human involvement in weapons which identify, select and attack targets', and that the UK 'does not possess fully autonomous weapon systems and has no intention of developing them'.[12] Whether an uncrewed vessel is even a warship in law, given that the Law of the Sea defines one as 'manned by a crew which is under regular armed forces discipline', is a question the Navy's own legal advisers have posed and nobody has yet answered.[13] So the model seems to stop at the weapon, which supports everything this series has said about how Float and Move are separable from Fight. The parts of the Hybrid Navy that can be bought as a service are the parts that do not fight, whilst the part that fights must have a human in the decision and, on the law as it stands, a crew under naval discipline somewhere behind it.
Private finance for defence capability is not new, although the UK has previously tried and stepped back from it. The Treasury announced in 2018 that it would 'no longer use PF2, the current model of Private Finance Initiative'.[14] The defence record behind that decision was mixed, where the National Audit Office found value for money was generally achieved across the MOD's PFI portfolio, but that the MOD 'does not always have the robust data necessary to understand the risks it is asking the private sector to bear'.[15] Voyager's 27-year tanker contract could not be shown to be value for money after four years of non-competitive negotiation, and the flying training system paid only a fraction of its early years' charges against the training it was designed to deliver.[16]
The question of private finance in Defence is now being re-opened from three directions:
fiscal pressure;
technology that arrives as a service from its manufacturer; and
a Government that says 'crowding-in private investment is vital'.
The newly established Defence Investors Advisory Group is 'exploring how private investment can be leveraged to help build the defence capability Britain needs'.[17] Uncrewed systems are an obvious candidate, although nothing published yet describes how such a service would be contracted. When the Navy needed a seabed-warfare mothership that would provide assured coverage, it bought one, judging that adapting a commercial vessel 'offered the best value for taxpayers and timely delivery'.[18] Subsequently, Project Lily, a £27 million contractor-supported autonomous survey service, was cancelled in January 2026 'following a decision to consider alternative internal options'.[19] [20] The lesson that carries from 2018 into the present debate is the NAO's finding that you cannot price a service you cannot measure.
Designed-in at contract
Contracting for capability has always been part of the support acquisition 'staircase', but one not generally seen in active use.[21] Instead, the MOD has tended to favour preservation of critical roles such as acting as the Design Authority and Integrator for the platforms it has acquired, despite the increasing reliance on industry to be able to do so. Whilst this arguably enables the overall safety case and operational risk to be managed within governmental control, it does introduce complexity at each of the design and sustainment intersections.
I have sat in rooms where a relatively simple change to a ship's system needed multiple organisations to agree before anyone could start: the manufacturer who owned the design data, the support contractor who held the maintenance record, the MOD team who owned the ship, the MOD team who were procuring the equipment, the Independent Safety Advisor, and so on. Nobody in the room was being obstructive and everybody wanted the change made, but delays came from the shape of the contracts, which had been written separately years earlier.
Aside from the obvious delays to the instantiation of capability, what is less evident is that acquisition and sustainment 'lock in' the delay, as a consequence of a fragmented supply chain and disconnected contractual arrangements with diverse, and sometimes competing, incentivisation.
Whose architecture, whose data?
The United States has had this argument in public, where its audit office found in 2025 that the Department of Defense 'can't maintain its weapon systems without intellectual property', that programmes 'don't always properly plan to get and manage the data rights they need', and that being tied to a single vendor was lengthening repair times across aircraft, ships, vehicles and submarines.[22] The galley ovens on USS Gerald R. Ford that sailors were contractually barred from fixing became the public face of the argument, and the Navy Secretary's 'I am a huge supporter of right to repair' the political one. In a separate case, a single contractor repair visit to a Littoral Combat Ship in Singapore was put at USD$1.2 million.[23] The right-to-repair provisions were stripped from two successive defence bills after industry objected,[24] and Congress is now considering a simpler rule where the government gets the rights it needs to maintain what it buys, unless the contractor can show why not.[23] Open architecture, the same audit office found separately, had been adopted on most programmes without anyone analysing the 70 per cent of life-cycle cost that sits in sustainment.[25]
In the UK, the 2021 industrial strategy committed to 'introducing Intellectual Property (IP) strategies into the MOD's acquisition processes' and to choosing open or closed architectures case by case, in a market where single-source contracts were about £8 billion a year.[26] The 2025 strategy does not revisit that commitment and, as noted above, does not mention intellectual property at all.[1] The Ministry's own support strategy, quoted in the last article, admits that its data is often incomplete and that it must be more assertive over data 'held by third parties'.[27] The MOD has already shown it can turn that intent into a standard, where the Land Open Systems Architecture, published this September, exists to let other suppliers compete for upgrades to Army vehicles, and says it 'helps reduce vendor lock-in'.[28]
The maritime equivalent so far is NavyPODS, where the Navy identified 18 potential categories of POD before it had a common standard to develop, field and support them, and has so far fielded little beyond the medical POD.[29] The Uncrewed Systems Centre, opened at Swindon in June 2026 with £310 million behind it, is described in the Defence Investment Plan as 'Defence's central authority for enabling and cohering the Uncrewed Systems ecosystem through the establishment and management of common standards, open architectures, data frameworks, and assurance models', and as the centre of excellence for uncrewed systems of systems.[30] That is a clear remit for the architecture, although the published documents do not say who leads it, how it reaches into the maritime domain, or, most important for this article, how this might impact the support systems that run on the architectures it sets.
The choice…
The clear choice is not a debate about old prime versus new prime, particularly given that either type of prime can lock support in. The traditional primes have held most of the support market for decades, much of it through long single-source arrangements, and both existing and new primes can readily open support up. Looking at the neo-primes, the route into the UK market runs through the two things the Industrial Strategy asks for: open interfaces and partnership with British suppliers. A US neo-prime that offers both is more attractive to the MOD than one that offers a closed system with a bundled support tail, and a traditional prime that offers the customer its own support data is more attractive than one that offers availability on terms only it can measure.
The choice is actually between support designed around the supplier's architecture and support designed around the customer's data. The first is faster to buy and simpler to hold to account, whilst the second is slower, and needs a customer that knows what data it wants and has the people to use it. It could also be argued that the latter is the only version compatible with a hybrid navy that is also supposed to be built by a broad UK industrial base. The MOD decides which it gets by what it writes into the contracts, which have too often been written separately and pulled in different directions.
That leaves the people who would have to use the data, on both sides of the contract, and the businesses that will have to change shape to supply it.
