The U.S. Navy and Coast Guard are buying into the same broad idea that unmanned aircraft can extend the reach of a force whose people, aircraft and attention are limited. But their priorities are sharply different.
The Navy’s MQ-20 experiment is about decision-making under pressure. The Coast Guard’s MQ-9B acquisition is about persistence staying over a vast ocean long enough to find a missing vessel, track a suspect contact or build a clearer picture of activity in the Arctic and Pacific.
One programme asks whether a drone can revise its plan. The other asks whether the United States can afford not to have an aircraft watching.
The Navy has tested adaptive-autonomy integration on the MQ-20 Avenger, a jet-powered unmanned aircraft used as a test platform. The concept is not that software replaces command. It is that the aircraft can react when the original plan no longer fits the conditions it encounters.
A drone may lose a communications link, encounter weather, detect a fuel constraint, receive updated threat information or find that a planned route is no longer available. In a conventional system, it may hold, return home or wait for a remote operator. In an adaptive system, it could assess approved alternatives and select a different route, altitude, sensor priority or return option.
The Navy’s central challenge is therefore not autonomy alone. It is explainability. A remote operator and mission commander must be able to understand why the aircraft changed the plan, what data it relied on, which authorised rules applied and whether the human can override it immediately. The system needs to produce more than an answer. It needs to show its reasoning in a form a trained operator can evaluate under time pressure.
The Coast Guard faces a more immediate problem that the sea is too large for the aircraft it has.
Its roughly $3 billion aviation recapitalisation, including four MQ-9B unmanned aircraft and six C-130J transports, pairs two very different forms of reach. The MQ-9B is intended to stay airborne for long periods, collecting imagery and sensor data across areas that would exhaust crewed patrol aircraft. The C-130J brings people, payload, speed, endurance and the ability to drop rescue equipment or move supplies after a hurricane.
The drone sees for longer. The C-130 can act. An MQ-9B can maintain a watch over a suspected smuggling route, fishing fleet, oil spill, ice field or vessel in distress. It can hand off a target from one shift to another and keep feeding imagery to command centres. But it cannot drop a life raft, carry a rescue team, transport emergency cargo or rapidly reposition a large crew and equipment package.
The C-130 can do those things, but it is a scarce, crewed asset. Every hour spent searching one patch of ocean is an hour unavailable for another mission.
That is why the two aircraft are complementary rather than interchangeable. The drone can narrow the search area, maintain surveillance and identify the contact that matters. The C-130 can bring aid, personnel and physical capability once the mission becomes urgent.
The difference is mission philosophy.
The Coast Guard is buying endurance to watch a busier ocean, where search and rescue, illegal fishing, migration, smuggling, storms and strategic competition increasingly overlap. Its drones are principally force multipliers: they stretch a limited human presence across more water.
The Navy is testing something more consequential. It is exploring whether an unmanned aircraft can become a more active partner in carrying out a commander’s intent when contact with the operator is delayed, degraded or lost.
The question is not whether drones will replace people.
It is whether institutions can build enough human oversight, technical assurance and operational discipline around them before the machines become fast enough to outrun the people responsible for them.