The founder's dilemma: American capital, British sovereignty, and the structure in between

Last week we laid out the trap. If you build in dual-use frontier technology and expand into the United States, you are not making one decision but three, governed by the national-security machinery of two governments pulling in opposite directions.

The first is a question of US ownership: how much control of your company you must hand to Washington to sell to it. The second is a question of UK call-in: whether your own government will let the deal proceed. The third is a question of export control: once your product turns partly American, who you are still allowed to sell it to.

We promised a harder question this week. If you want to sell to both the UK and US governments, and raise the late-stage US capital to do it, how do you build so that no single decision quietly forecloses the others?

The honest answer is that no single solution works in every case. The right structure for a Series A company with the UK Ministry of Defence as its first customer is not the right structure for a Series B company with a signed US government contract and a US lead investor. The variables are too many and they shift over time. What we offer instead, drawn from years of guiding founders through these decisions, is a simple framework that produces your answer rather than a generic one, and holds up as circumstances change.

The framework comprises three lenses and two dials. Whenever one of these decisions lands in front of you, examine it through three lenses in turn: what your end customers demand of your structure, what your investors will require in return for their cheque, and what the tax and legal position can actually bear. Then turn two dials: your stage, because an early-growth company still holds options a late-stage one has traded away, and your sector, because a purely commercial DeepTech business with no defence or dual-use dimension escapes almost all of this.

We will apply the framework to the three pressure points we identified last week, in turn.


Governance: how much of your company answers to Washington?

The first pressure point is governance, and it begins with a fact that surprises many founders: a British-owned company can be awarded US defence contracts. Foreign ownership is not a locked door. What it does is place you inside a US system designed to manage the perceived risk of that ownership, and the real question is how much of your company that system ends up controlling.

Two factors determine the answer, and the discipline is to keep them separate. The first is the sensitivity of the work: the most secret programmes demand the heaviest oversight no matter how you are owned, while less sensitive contracts ask far less. The second is how foreign your ownership looks, and this is where the investor lens leads, because the Delaware flip we examined last week changes the picture directly. The catch, which runs through this entire series, is that the same flip eases the problem by making you American.

The sensitivity spectrum runs from gentle to severe. For low-sensitivity contracts, you fence off classified information inside the company and otherwise run the business much as before. For the most secret programmes, the US business must be governed by a board of security-cleared American citizens, nominated through a US-approved process; you, the foreign owner, are barred from that board and cannot direct the people on it. You keep your shares and your profits but lose your grip on how that part of the company is run. This is why how tightly you draw the line around the sensitive work matters so much: concentrate it in a defined, ring-fenced US unit rather than letting it spread across the company, and you confine the heaviest controls to that unit while the rest of the business stays in your hands.

Ownership is the second factor, and this is where the flip does its work. At the same level of sensitivity, a wholly foreign-owned company faces the more intrusive instruments, because the regime is calibrated to how much foreign control actually exists. Place a US holding company above the business and bring American investors onto the register, and you move towards the lighter end of the same spectrum. The flip does not switch the scrutiny off, since the US authorities weigh country of origin and sensitivity regardless. But lowering the foreign-ownership factor is often enough to keep you clear of the heaviest instruments, letting you retain far more control than a foreign parent otherwise could.

Put the two together and the shape of the governance question is clear: the sensitivity of your work sets the ceiling on what Washington will demand, and how foreign you look sets where beneath it you land. Neither is fully within your gift, but both can be shaped, by scoping the sensitive work narrowly and by weighing when, and whether, to flip. One thing to watch is a proposed US rule that would, for the first time, pull unclassified defence work into this same system: not yet in force, but a clear sign of the direction of travel. And your UK institutional investors will examine all of it closely, because every piece of control you concede to Washington is control they no longer hold.


Block or unwind: the move that pleases Washington can alarm Whitehall

Here’s the shift that catches founders off guard. The very flip that eased your American problem can trigger a British one, because moving your company's ownership and its underlying technology to a US holding company is precisely the kind of transaction your own government has given itself the power to stop.

This is where the tax and legal lens leads, through the National Security and Investment Act. It lets the British government review a deal on national-security grounds, then clear it, attach conditions, block it, or unwind it after completion. Two features matter directly. The trigger is lower than founders expect. Notification becomes mandatory not just on a change of control, but when any single holding crosses 25% of shares or voting rights, so a sizeable minority round, or the flip itself, can be the event that brings you into scope. The consequence of getting it wrong: a deal that should have been notified but was not is legally void, with civil and criminal exposure attached. This is not merely a risk of delay. It is the risk that the centrepiece of your fundraise is undone by operation of law.

The reach is what surprises people. The Act applies not only to a foreigner buying into a British company but to the transfer of sensitive technology out of British control, exactly what a flip to a US parent accomplishes. The regime currently covers 17 sensitive sectors and is tightening: a reform confirmed earlier this year will carve out semiconductors as a category in its own right, part of a clear pattern of the rules closing in around the very technologies this series is about.

None of this argues against flipping. It argues for sequencing. The same share-for-share exchange can usually be structured to avoid an immediate UK stamp duty charge, but the relief is granted narrowly and easily lost if your later investment plans are not thought through at the outset. The flip is not administrative housekeeping to tidy up after the round closes. It is a decision with two governments watching, and it rewards the founder who plans it rather than processes it.


Sell-back: who you can sell to once the technology is part-American

Suppose you navigate the governance question and the British review, take the US money, and win the US customer. A quieter cost surfaces later, and it falls under the end-customer lens, because it decides who you are still allowed to sell to.

The principle to hold onto is that control follows content, not location. It does not matter that your engineering team sits in Bristol: once your product contains American technology of the controlled kind, your freedom to sell it on runs through Washington. Defence-grade American content can carry that constraint in any quantity; dual-use American technology carries it once it passes a threshold proportion of the whole. And the constraint bites in the direction founders least expect: selling the finished product back to the United Kingdom, including to the Ministry of Defence that may have been your first customer, can require American permission.

There is a mechanism that eases this among allies, worth understanding precisely because it is easy to over-read. A reciprocal arrangement finalised at the end of last year lets vetted British and American entities trade a wide range of defence items without seeking individual licences each time. But it is a faster lane, not an exit from the road. The underlying American control over the technology remains, and membership of the trusted community is itself conditional. The administrative friction falls. The dependency does not.

For the founder, the end-customer lens turns this into a question to ask early rather than discover late. If selling to the British state is central to what you are building, every increment of American controlled content you absorb hands a third government a say over a customer relationship you depend on. That may be a price worth paying for the capability the American technology brings. It may not. The error is to find out only at the point of sale.


The dial in motion: one founder, two different answers

The framework earns its keep when you watch the dials turn, so take one founder at two moments in the company's life and see how the same questions resolve differently:

Approaching Series A, selling mainly to British customers with the Ministry of Defence first among them, the sensible posture is restraint. The flip can wait: no American investor is yet demanding it, and holding it back keeps your options open rather than committing you to the US path before you have to. The governance question barely bites, because the work has not yet reached the sensitivity that summons the heavier controls. The binding constraint is the sell-back one: keep American controlled content low while you still can, and you preserve the freedom to choose later. This is the watching-brief stage, where the disciplines matter precisely because none of them is yet forcing your hand.

Approaching Series B, the dials have turned. There is a US government contract in hand and an American lead investor at the table, and the flip is now the price of the round rather than a deferred option. The questions invert. Governance moves to the foreground: even with the group now sitting under a US parent, the most sensitive defence work still has to be walled off inside a cleared, security-vetted entity, and the design task is to scope that entity tightly so the heavy controls fall on it alone rather than reaching across the whole company. The British review must be sequenced, not stumbled into. And the sell-back constraint has hardened from a thing to monitor into a thing to manage.

Same company, same founder, a wholly different answer, because stage and sector moved. The lesson is not that one path is right, but that the founder who knows where the dials are set can build on purpose, while the one who does not will find the structure chosen for them, one reasonable round at a time. It is the same logic that ran through our argument about how a board re-forms at every round: the configuration you finally understand is the one about to change, and the decisions that bind you hardest are the ones you let happen by default.


The choice beneath all three

Step back from the cap table and the deeper pattern shows itself. These three pressures bite at all because Britain builds frontier companies well but funds their scaling badly. Strong at Seed and Series A, thin at the growth stage, the UK pushes its most ambitious founders towards the one source of capital large enough to scale them: American. That dependency on a single source of capital is what hands Washington the leverage every one of these regimes expresses.

And a dependency is only as durable as the politics holding it together. Consider the headline arrangement meant to cement the transatlantic technology relationship. Signed with great ceremony last autumn, the Technology Prosperity Deal was paused within three months, not over anything the technology companies had done, but because the two governments fell out over unrelated disputes on taxation and online regulation. An arrangement founders were relying on to underwrite their US expansion was suspended by forces entirely outside their control. It is the same lesson Part 1 drew from a single company losing access to an AI model overnight, now playing out at the level of nations.

So, the question to sit with is larger than any one financing. Structuring deliberately, taking these three decisions consciously rather than letting them be made for you, is the founder's best available defence. But it is only a defence, not a cure, because the root problem is national: Britain has no pool of capital large enough to scale its frontier companies without turning to the United States, which is exactly why these decisions are being forced on founders now.

Until these circumstances change, the American path will keep arriving bundled with another government's priorities, and every founder in dual-use technology will face the same three decisions with the same narrowing room to manoeuvre. The task, for now, is to see all three coming, and to build so that saying yes to one never quietly forecloses the rest.

 
Let's talk.