The Yards Are Korean. The Question Is Whose Navy.

Week 41 | September 2026

We start with three facts from this year.

Hanwha, a South Korean company, owns a shipyard in Philadelphia. In June, the US Navy (Navy) sent requests for information to three Korean builders, asking what they could produce and how fast. And on 6 July 2026, Canada named Germany's TKMS as preferred supplier for its next submarine fleet, after a year weighing the faster, cheaper Korean bid.

On their own, these are: a shipping note, a procurement note, and a defence-policy brief. Together, they add up to one thing: the capacity to build the alliance's warships increasingly sits outside the alliance's own industrial base.

Back in Week 21 I called this "the allies are shopping, not building." True then, true now. But shopping assumes you can complete the purchase. This brief is about the part of the transaction that doesn't clear, and why.

Everyone who writes about foreign shipbuilders and the US Navy reaches for the same statute. The Byrnes-Tollefson Amendment, now 10 U.S.C. § 8679 (a), says: except as provided in subsection (b), no vessel to be constructed for any of the armed forces, and no major component of the hull or superstructure of any such vessel, may be constructed in a foreign shipyard. Six decades old and ironclad, apparently.

Read that again. § 8679 (a) bars construction in a foreign shipyard. That means one on foreign soil. It's silent on who owns the yard.

Hanwha owns Philly Shipyard outright: a $100 million deal, CFIUS-cleared in September 2024, closing that December. American soil, American workers. So the commonly cited decades-old ban doesn't apply to it. A Korean-owned yard could, on the plain text of § 8679 (a), lay down grey-hull naval tonnage in Pennsylvania tomorrow.

It's already partway there. Philly is turning out MARAD training ships and ten commercial tankers, and in March Hanwha's US arm won a subcontract, under Vard Marine, on the Navy's next light replenishment oiler design, the first time a Korean firm has entered a US Navy shipbuilding program at the design stage. The Jones Act doesn't block it either: that law is about the ships and their owners, not the nationality of the yard.

So if the law isn't the wall, what is?

The Navy’s June requests for information went to HD Hyundai and Hanwha Ocean for destroyer-sized combatants, with Samsung added for oilers. The trigger was about as subtle as they come: Trump reportedly turned to President Lee at the G7 and asked whether Korea could quickly build ten American warships.

Here's what an RFI is, technically: Market research. Under federal acquisition rules it gathers capacity, pricing and timelines to shape a future buy. It commits the government to nothing.

Now look at what it measures: whether Korean yards can build warships fast. That was never the open question. Korea's commercial output settled it years ago; its yards rank among the world's largest by tonnage. The RFI interrogates the one variable nobody doubted, and steps politely around the one that actually governs the outcome.

Trump nudged that outcome in August with a memo expanding the White House's "Finland Model," using the statute's own national-security waiver rather than repealing it: a qualifying foreign builder can build its first couple of hulls at home before shifting the rest to a US yard it owns. Sounds like the door opening. Then look at which ships walk through: sealift and tankers, plus a surface combatant nobody has designed yet. The low-classification end of the fleet. That selection is the tell.

Here's the machinery that matters, and it's genuinely dull, which is why it often gets skipped.

When a foreign company owns a US defence facility, the Pentagon doesn't hand it classified work on trust. It runs the yard through a regime called Foreign Ownership, Control or Influence, or FOCI. A foreign-owned yard can hold a facility clearance, but majority-foreign ownership, like Hanwha's, means a Special Security Agreement. And once the work touches the classified guts of a combat system, even that isn't enough: it needs a separate National Interest Determination, category by category, signed off by the agency that owns each secret.

There's a shortcut. In 2019 Congress waived that determination for companies owned from inside the alliance's inner industrial circle, the National Technology and Industrial Base: the US, the UK, Australia and Canada.

Korea isn't in it.

So a Hanwha yard in Philadelphia can weld tankers and auxiliary hulls all day. A warship's classified fit-out is a different gate, taken category by category. The yard has the steel and the welders. What it lacks is the clearance.

If that sounds abstract, watch it decide two real competitions.

Start in Korea's own backyard, the cleanest example there is. This year's KDDX destroyer contract came down to HD Hyundai against Hanwha. Hyundai scored higher on technical merit. Then a 1.2-point security penalty landed: a group of Hyundai employees had been convicted of photographing and leaking Hanwha's classified KDDX design material. It flipped the result, and Hanwha won by roughly half a point. The more capable shipbuilder lost the destroyer on security governance, inside its own country, where nationality wasn't even in question.

Then Canada. Ottawa spent a year evaluating Korea's KSS-III submarine, faster and cheaper, then on 6 July went with Germany's TKMS instead. Carney's reason: interoperability with NATO partners already sailing the boat. Hanwha sits as reserve. Capability was on the table. It lost to fit.

Even the establishment concedes it. Matthew Paxton, president of the Shipbuilders Council of America, warned that allied shipyards are no seamless substitute, because warship work runs on classified combat integration at the edge of what's shareable even with close allies. 

Put the three side by side and the pattern holds: in each, capability was the easy part, and clearance or alliance fit was the deciding one. Capability is abundant, increasingly available, and in Philly's case entirely legal. What's scarce is an acquisition-and-clearance architecture built to absorb it from where it now lives.

Washington can't decide whether it wants to use its own workaround. The August memo leans in. The FY2027 defence bill grinding through Congress leans hard the other way: an amendment barring funds for any battle-force ship built in a foreign yard, while the White House's own budget office opposes the section that contradicts the President's plan. That fight isn't settled. It's the fight that decides how much of this actually happens.

Korea can build the ships. Whether the alliance's clearance architecture will let it is the question the RFI can't answer, and the question Byrnes-Tollefson was never written to ask.

The world's navies are all fielding uncrewed underwater vehicles. Almost none of them has said what happens when the one in its own waters isn't theirs. No flag, no crew, no master to hail. Next week: whether detection even works against small electric UUVs, who you're allowed to blame when there's no crew to blame, and whether any navy has actually published a rule for engaging one.

Since you have been, thanks for reading.

Cheers,

Mick

Ocean Tech Intelligence provides informational analysis only. Nothing in this publication constitutes financial, investment, legal, or strategic advice. Readers act on this content at their own risk. For full details see our Disclaimer.