By age twelve, I could don, clear, seal, and check an M17 protective mask in under nine seconds – a drill my father ran on the floor of our quarters at Camp Humphreys, where he served as garrison commander, until it stopped being something my twin sister and I thought about and just became something we did.
I went back to Korea years later in uniform to help run Ulchi Freedom Guardian, the exercise that became Ulchi Freedom Shield, and returned again working on illicit finance and fintech issues with Korean counterparts.
Among three very different reasons to be there, I learned the same thing: this alliance was never kept alive by declaring it, but by using it, on a schedule, whether or not that week felt like it called for it. After all, we weren’t going to get to choose in advance when we’d need it.
That’s what makes President Trump’s 16 August last-minute order directing Secretary Hegseth to “substantially reduce” this year’s Ulchi Freedom Shield a surprise to many who have served in or supported that Indo-Pacific theater.
The order landed hours before the eleven-day exercise was set to begin, with roughly 18,000 South Korean troops participating alongside American forces. Trump cited cost, a desire not to provoke Pyongyang, his rapport with a leader the United Nations has found responsible for crimes against humanity, and Seoul’s refusal to help on Iran. The exercise ended early and significantly downsized the combined field training components.
Ulchi Freedom Shield runs through the Combined Forces Command pulling in US Forces, Korea, United Nations Command, and the ROK Joint Chiefs of Staff to rehearse and build integrated force functionality under wartime conditions. These kinds of maneuvers either work under pressure or not, with no way to find out except by doing it. A failure caught mid-exercise gets fixed before it matters, but caught in a war, it gets people killed.
Exercises change, and they should change when commanders determine that a different design will better prepare the force for the contingency at hand. The harder question is what happens when an exercise becomes leverage in a broader political bargain. Alliances are instruments of statecraft, and Washington should use them to pursue its interests. But leverage that degrades the military capabilities, command relationships, or confidence on which the alliance depends can consume the very asset it is meant to strengthen.
The Iran linkage doesn’t hold up well even on its own terms. In announcing the reduction, the administration listed Seoul’s answer on Iran as a reason while calling the connection “somewhat unrelated” in the same breath.
It also doesn’t square with how South Korea’s military is actually built. Its roughly 450,000-strong force runs almost entirely on conscription, structured around a single legal mandate: territorial defense of the peninsula, not expeditionary power projection.
That’s a different kind of military than the ones that show up in coalition operations far from home as a matter of course.
And yet this is also a country that sent more than 300,000 troops to Vietnam and deployed to Iraq and Afghanistan, none of them fighting for Korean survival, either. Whatever explains Seoul’s (widely shared) caution on Iran, a general reluctance to stand alongside the United States isn’t it.
What’s missing is any strategic case for why that one answer needed to shape how many troops rehearsed defending the DMZ eleven days later. If Washington wants greater Korean support beyond the peninsula, it has more targeted tools for demanding it than weakening the machinery built to defend Korea itself.
South Korea anchors American deterrence against a nuclear-armed adversary and one of the largest concentrations of artillery within range of a major capital anywhere on earth, under a mutual defense commitment dating to 1953.
It’s also one of the two indispensable American alliances in northeast Asia, alongside Japan, and together they give Washington something Beijing can’t match on its own.
Victor Cha’s Senate testimony on Indo-Pacific alliances puts roughly 88,500 American troops forward-based in Korea and Japan combined, against a Chinese overseas basing network that, even now expanding to a second facility in Cambodia, remains a fraction of what decades of alliances have built for Washington.
That is the strategic advantage the alliance provides: not simply access to territory (which can also be critical for warfighting and resupply), but a network of capable allies already integrated into the region’s security architecture.
And it’s why Kurt Campbell, who ran Indo-Pacific policy at the NSC and State Department, has called an “America First” approach to the region a “recipe for failure” against an assertive Beijing, precisely because it weakens alliances like this one.
Getting Korea and Japan to operate within something closer to a mutual security environment, rather than two bilateral relationships that happen to sit near each other, took years of diplomatic effort to unwind decades of antagonism between Seoul and Tokyo.
The 2023 Camp David summit turned that opening into something durable, and by December of that year the three governments had activated a real-time North Korean missile warning data-sharing system and locked in a multi-year trilateral exercise plan. Korea and Japan aren’t interchangeable, but they’ve become complementary pieces of one deterrence structure. Weakening one doesn’t leave the other to pick up the slack. It leaves the whole network thinner.
And the alliance has become much more than just a conventional military relationship. The Nuclear Consultative Group, created by the 2023 Washington Declaration to institutionalize U.S.-ROK nuclear consultation, held its sixth meeting this past June. Korea also sits inside the Multilateral Sanctions Monitoring Team Washington helped stand up to track North Korea’s crypto-laundering – proceeds that, per a recent Bloomberg Economics analysis, may have handed Kim’s regime as much as $22 billion in foreign revenue since 2022.
Industrial investment is growing as well. Hanwha has put $5 billion into its Philadelphia shipyard, HD Hyundai just landed the Navy’s first Korean-yard MRO contract, and Hanwha has bid roughly $1.2 billion for Austal USA, whose Mobile yard builds Navy and Coast Guard ships and submarine modules for General Dynamics Electric Boat. Washington is asking Seoul to invest deeper into the architecture of American security while making the foundational commitment underneath it less predictable.
Allies elsewhere are already pricing this in. A May 2026 European Council on Foreign Relations survey found that only 11 percent of Europeans now consider the United States an ally, down from 22 percent at the end of 2024, with majorities in every country surveyed doubting Washington would defend them. Korea is now facing a version of that same recalculation, on an alliance with more riding on American follow-through than almost any other in the world.
My father made my sister and me practice putting on a gas mask at Camp Humphreys because he understood a simple principle: preparation is most valuable before anyone knows whether it will be needed.
The U.S.-ROK alliance is valuable because Washington and Seoul have spent seven decades building command relationships, capabilities, institutions, and habits that would allow them to act together in a crisis. Washington has spent years extending that architecture to Japan and into nuclear, cyber, financial, and industrial cooperation.
But the recent pattern is moving in the opposite direction. Washington is asking allies to shoulder more of the burden of European and Indo-Pacific security while making the commitments they plan around less predictable – asking South Korea to invest billions in American shipbuilding while making the military relationship more conditional, building trilateral capabilities with Seoul and Tokyo while weakening one of the exercises that makes those capabilities credible, asking allies to help enforce sanctions and expand defense production while signaling that a disagreement in one theater can affect the security commitments at the heart of another.
The United States should ask more of its allies. But Korea already is: billions committed to American shipyards, a seat inside the sanctions architecture tracking North Korean crypto theft, decades of troops sent into wars that were never its own to fight. Asking allies to keep investing in a common security system requires giving them confidence the system itself will remain reliable.
That does not mean every exercise must remain unchanged, or that alliances should be immune from hard bargaining. It means the United States should be careful about bargaining away the very capabilities that make the alliance itself credible. Exercises, command relationships, intelligence sharing and the institutions built around them are not just diplomatic favors; they set the infrastructure that allows Washington and its allies to act together when deterrence fails.
The question raised by Ulchi Freedom Shield is bigger than this year’s exercise; it’s whether Washington still understands that the infrastructure of deterrence has to be maintained before the crisis arrives rather than trying to reconstitute after it does. My father knew there would be no time to learn how to put on the mask once the siren sounded. Neither should the United States expect to rediscover the strength of its alliances after it needs them.
Carole House is a former White House National Security Council Special Advisor, co-author of President Biden’s executive orders on digital assets and cyber innovation, former Chair of the CFTC Technology Advisory Committee, member of Governor Newsom’s Innovation Council, and Senior Fellow at the Atlantic Council GeoEconomics Center.
