A Torpedo From an Uncrewed Submarine Is Not Yet a Capability

1 The Test And Its Exact Claim

The Royal Navy launched a U.S. Navy Mk 48 heavyweight torpedo from a British Excalibur uncrewed underwater vehicle on September 13 at the British Underwater Test and Evaluation Centre in Scotland. The Navy's release stated that the event validated the mechanical, electrical and software integration of a U.S. kinetic payload onto an allied autonomous vessel, and that it proved allied forces can rapidly achieve weapon interchangeability. Adm. Daryl Caudle, the Chief of Naval Operations, called it a historic first for allied undersea warfare and framed it around integrating weapons, payloads and mission architectures across allied platforms.

The claim is specific and worth taking at face value: integration was validated. The claim is not that an uncrewed submarine conducted a patrol and engaged a target.

2 Interchangeability Under AUKUS Pillar 2

The launch sits under AUKUS Pillar 2, the trilateral arrangement between the United States, United Kingdom and Australia that is meant to enable uncrewed undersea vehicle payload systems usable across each country's UUVs, per a May 2026 agreement. Interchangeability is the meaningful concept here, because it means a weapon developed by one member can be carried by another member's platform. The Mk 48's current variant, the MOD 7, is itself the product of a joint development program with the Royal Australian Navy, so the torpedo in the water was already a trilateral object.

3 The Torpedo That Does Not Fit

A detail from the event sharpens the picture. The Mk 48 is roughly 19 feet long with a 21-inch diameter. Excalibur, the Royal Navy's first uncrewed submarine, unveiled in May 2025, has no torpedo tubes but a large central payload bay with doors top and bottom - and, per the UK outlet Navy Lookout, the Mk 48 is too long to fit in that standard bay. Video of the event indicates the torpedo was carried under the roughly 40-foot submersible and released from a cradle attached to the hull.

An external cradle works for a trial. It is a poor arrangement for a deployed vehicle, because it changes the hydrodynamic and acoustic signature of the host, complicates handling in port, and raises questions about safe carriage and release in a seaway. The interchangeability demonstrated is real at the level of interfaces and software; the stowage problem it exposed is equally real.

4 Reload And Resupply Are The Operational Question

An attack platform is defined by its magazine depth and its reload cycle. A UUV that carries one externally mounted torpedo has a capability measured in single shots, and the operational tempo it can sustain depends on how quickly it can be reloaded, where, and by whom. No public reporting from the September event addresses reload. That is not a criticism of the test; it is a marker of how much of the capability remains undefined.

5 Authority To Release

Separate from engineering is the question of who authorizes a weapon release from a vehicle with no one aboard. The trial did not involve an engagement, so the question went unaddressed. For an uncrewed submarine operating in a communications-denied undersea environment, the release authority question is not hypothetical: it determines whether the vehicle can be employed at all in the conditions where it would be most useful.

6 The Follow-On And The Distance

The U.S. and Royal Navies are continuing with Project CUTLASS, described as an acceleration of the work. That is the path from a validated interface to something operationally relevant. The distance is set by the unaddressed items: stowage, reload, authority, and a concept of operations that says what an uncrewed undersea vehicle is for and how many are needed.

7 The Bottom Line

Launching a U.S. heavyweight torpedo from an allied autonomous submersible is a genuine first and a real step toward weapon interchangeability among AUKUS partners. It is also a proof of concept, by the Navy's own description, and the word matters. Between a released torpedo and a fielded uncrewed attack capability sit stowage, reload, release authority and doctrine - none of which the September test claimed to solve.