The Coming Ghost Fleet
NAVY APPEARS TO SHIFT FROM OPTIONALLY MANNED TO FULLY AUTONOMOUS WARSHIPS
BY DAN TAYLOR
The U.S. Navy’s vision for uncrewed
warfare may be getting lonelier. After years of exploring
“optionally manned” designs that could accommodate
Sailors when needed, the service may be leaning
toward a future where warships are built without
accommodations for human crews — no berthing, no
galleys and no heads.
The shift represents more than an engineering choice;
it’s a fundamental reimagining of naval power projection,
where autonomous vessels operate for months at sea
without anyone aboard. The question is whether removing
humans entirely delivers the cost savings and tactical
advantages the Navy expects, or creates vulnerabilities the
service hasn’t fully reckoned with.
An example of where the Navy may be headed is the
USX-1 Defiant, an uncrewed vessel DARPA christened
at Everett Ship Repair in Washington this summer. The
180-foot vessel was designed from the keel up to never
accommodate humans aboard.
“Defiant is a tough little ship and defies the idea that we
cannot make a ship that can operate in the challenging
environment of the open ocean without people to operate
her,” Greg Avicola, program manager for DARPA’s
No Manning Required Ship program, said during the
christening ceremony.
At 240 metric tons with a simplified hull design, Defiant
can be built and maintained at nearly any port facility
DARPA’s No Manning Required Ship (NOMARS) program has built a demonstrator ship
designed to operate autonomously for long durations at sea, the USX-1 Defiant.
or Tier III shipyard that traditionally supports yachts,
tugs and workboats, not just the handful of major naval
shipyards that build destroyers and aircraft carriers. The
vessel is designed for extended ocean voyages, capable
of operating in sea state five with no degradation and
surviving much higher seas.
“She’s no wider than she must be to fit the largest piece
of hardware and we have no human passageways to
worry about,” Avicola said.
After completing at-sea demonstrations, Defiant will
transfer to the Navy’s Unmanned Maritime Systems
Program OfÏce, becoming the service’s first fully
autonomous medium uncrewed surface vessel. That
transition matters because it moves the concept from
experimental prototype to potential production model —
a vessel that could be replicated across the fleet.
DARPA Director Stephen Winchell framed the capability
in strategic terms during the ceremony: “Defiant
class vessels provide cost-effective, survivable,
manufacturable, maintainable, long-range, autonomous,
and distributed platforms, which will create future naval
lethality, sensing and logistics,” he said.
Congress appears convinced of the concept’s potential.
The July reconciliation bill appropriated $2.1 billion “for
development, procurement, and integration of purpose
built medium unmanned surface vessels.”
Questions Remain
Yet some analysts have identified questions about how
autonomous vessels fit into the Navy’s operational plans,
and whether the service is moving too quickly toward a
concept it hasn’t fully tested.
A March 2025 Congressional Research Service report
examining the Navy’s unmanned surface and undersea
vehicle programs notes that while the service envisions
medium uncrewed surface vessels (MUSVs) supporting
intelligence, surveillance, reconnaissance and targeting
missions, the path from prototype to operational
fleet remains unclear. It notes the Navy’s fiscal 2025
budget submission doesn’t include the procurement of
operational MUSVs over the next five years, and also
questions whether the Navy has fully developed the
concept of operations for such vessels.
CRS also asked how the Navy will manage these vessels
during port visits or in situations requiring diplomatic
judgment, as well as what happens when autonomous
systems encounter scenarios their programming never
anticipated.
The CRS report notes the Navy is building institutional
capacity alongside hardware, establishing career paths
for unmanned surface warfare officers and training
enlisted robotics specialists at Carnegie Mellon
University. But organizational structure and training
pipelines don’t answer the operational questions about
how a hybrid fleet of crewed and uncrewed vessels will
actually, function together in contested waters.
“The issue for Congress,” the report concludes, is
“whether to approve, reject, or modify the Navy’s
acquisition strategies and funding requests for these
large UVs,” given that the proposals “pose a number of
oversight issues for Congress” around technical risk,
cost growth and operational concepts that remain works
in progress.
The economic logic behind fully uncrewed designs
is straightforward, said Brad Martin, senior policy
researcher at the RAND Corporation. Building ships
that might occasionally carry crews means designing in
systems that add cost even when unused.
“You presumably save a little bit in your manning cost
with an optionally manned vessel,” Martin said. “But
it’s still in the design of the ship. You’ve got to put in
refrigeration, air conditioning, water and galleys and
berthing and all that stuff. So, it can be expensive.”
But eliminating human presence entirely means
accepting operational limitations. Martin pointed to
scenarios requiring nuanced judgment — boarding
operations, responses to diplomatic incidents or
situations where rules of engagement remain ambiguous.
“There’ll be times when you might feel better having
a person there because the situation’s confused, or
whatever it is,” Martin said. “Unmanned’s got to be
controlled by a human one way or the other. And I don’t
think that that’s infallible.”
For missions the Navy has prioritized, primarily missile
carrying magazine ships and ISR platforms, those
scenarios may be limited. Martin suggested risks can
be mitigated through remote oversight when vessels
are already being monitored for their data collection
capabilities.
“If you’re that worried about it, you can increase the
ability to intervene remotely,” he said. “If it’s something
where you’re already monitoring it because of its data
capability, you can mitigate the lack of human judgment
in a crisis situation.”
The Navy’s apparent pivot toward fully uncrewed vessels
seems increasingly inevitable, driven by cost pressures,
strategic requirements for distributed operations, and
just the general shift throughout human society toward
using autonomous assets. With billions appropriated and
prototypes like Defiant demonstrating technical viability,
the question is no longer whether ghost ships will join the
fleet, but how many and how soon.
Yet as the service enters this new era, fundamental
questions about operational concepts, autonomous
decision-making, and hybrid fleet integration remain
unanswered — questions that will shape whether
uncrewed platforms deliver the revolutionary capability
the Navy envisions or expose vulnerabilities in ways the
service hasn’t fully anticipated.
Dan Taylor has covered the U.S. Navy and the Pentagon since 2007 for a wide range of publications,
focusing particularly on Pentagon acquisition and the latest in defense technology. Currently, he is
technology editor for Military Embedded Systems