The U.S. Navy’s unmanned surface vessels (USVs) are proliferating across multiple areas of operation, participating in increasingly more elaborate tests, integrating with partner navies, and being tactically employed. However, recent incidents involving the seizure of USVs highlight an ambiguous legal environment that foreign adversaries can exploit in the gray zone (described by former Deputy Secretary of Defense Robert Work as avenues of approach in which adversaries use agents, paramilitary organizations, deception, infiltration, and persistent denial to achieve their goals).
Charged with an ambitious goal of employing one hundred USVs by the end of 2023, the U.S. Fifth Fleet is at the forefront of incorporating technology and innovation in both unmanned and autonomous systems into real-world operations. The pressure for the Navy to field unmanned systems and exploit their efficiency is intense. However, the Navy must also protect these USVs from interference by adversaries’ gray zone actions. In this environment, legal wrangling and risk mitigation will constrain rather than expand USV employment. In other words, the Navy’s likely response to these gray zone actions might constrain the tactical use of USVs rather than expand it. The result would be a counter-revolution in USV tactical employment.
Recent seizures of Navy unmanned vessels by China and Iran showcase a distinction between the tactical employment of USVs and functional testing and experimentation. With these seizures, adversaries are claiming questionable justifications for why the seizures were acceptable, if not necessary—to ascertain their seaworthiness and ensure other vessels can operate safely in their vicinity, for example.
USV Development
In September 2021, Fifth Fleet established Task Force 59 to “rapidly integrate unmanned systems and artificial intelligence with maritime operations in the 5th Fleet area of operations,” the first task force of its kind. Task Force 59 has claimed several successes, including more than 15,000 total USV sailing hours and participation in six bilateral and multilateral exercises since its inception.
Meanwhile, Unmanned Surface Vessel Division 1 (USVDiv-1) leads testing and experimentation with larger, more robust USVs made up of the Sea Hunter and Seahawk, as well as two Ghost Fleet Overlord offshore support vessels designated as Nomad and Ranger. USVDiv-1 is a part of Surface Development Squadron (SurfDevRon) One, the experimental unit leading all three Zumwalt-class destroyers and other unmanned surface vessels. SurfDevRon One has its sights on developing high-end USV operations and technologies, as the Navy strives to reach its goal of an additional 150 unmanned vessels by 2045.
In the near term, Fifth Fleet’s ability to achieve its USV employment goal will depend not on technological advances or experimentation, but on whether Navy leaders can manage the risks associated with protecting USVs. In the long term, the United States intends to acquire larger USVs “as part of an effort to shift the Navy to a more distributed fleet architecture, meaning a mix of ships that spreads the Navy’s capabilities over an increased number of platforms and avoids concentrating a large portion of the fleet’s overall capability.” This broader goal is also in jeopardy unless the Navy can come to terms with how to protect unmanned ships. Navy leaders already recognize this and are suggesting that smaller USVs like the Saildrones seized by Iran must have no intrinsic value to adversaries. Resolving how the Navy will protect USVs with classified hardware remains to be seen.
Recent Case Studies and the Legal Ambiguity
In December 2016, a Chinese rescue and salvage ship seized one of two U.S. low-buoyancy underwater glider drones from the survey ship USNS Bowditch (T-AGS-62) that were surveying in waters claimed as economic exclusion zones (EEZs) by both China and the Philippines. The Chinese crew took the drone on board for inspection. Soon after, China released a statement explaining that the “confiscation was meant to prevent the device from posing danger to the safe navigation of passing ships and personnel.” In response, the U.S. government demanded the return of the device, which the Chinese eventually did.
More recently, an Iranian Revolutionary Guard Corps–Navy ship in the Arabian Gulf towed away a U.S. Navy Saildrone USV on 29 August 2022, and on 2 September an Iranian destroyer in the Red Sea picked up two Saildrone USVs and hauled them on board. In each case, the USVs were eventually returned, though installed cameras were missing from both vessels in the latter incident. Iranian officials claim they were acting “to prevent possible accidents,” “to prevent possible maritime incidents and safeguard shipping lanes,” or “supporting safe and secure shipping,” suggesting perhaps that the vessels’ “navigation systems had failed.”
Initial reports even indicate it is possible these Saildrones may have become disabled for some reason. Regardless of whether the USVs were disabled, future adversary vessels can cite safety of navigation to justify seizing USVs, a misleading and pretextual narrative to justify interference with USVs. Several important issues remain unsettled as a matter of international law. They range from the status of USVs as “ships” or “vessels,” whether USVs can claim sovereign immunity as government-owned vessels, and what level of jurisdiction nations can hold over them when in that nation’s EEZ.
Regarding the status of USVs as “ships” or “vessels,” on review of the United Nations Convention on the Law of the Sea (UNCLOS), many experts conclude that, “Although UNCLOS does not define the term ‘ships’, when reading the instrument in the context of its own text, as is appropriate pursuant to Article 31 of the Vienna Convention on the Law of Treaties, it appears to consider that ships are manned.” Similarly, the International Convention for the Safety of Life at Sea (SOLAS) stipulates that immediate human control of a ship’s steering should be established in areas of high traffic density. It is unclear how this might apply to USVs when a remote-control mode is a possibility, but clearly compliance would be problematic when no remote-control option is available.
Sovereign immunity for USVs as warships or government-owned vessels is also a legal gray area. The Navy argues that a drone or USV can be extended the same rights barring “arrest” of warships, but UNCLOS defines a warship as a “ship belonging to the armed forces of a State bearing the external marks distinguishing such ships of its nationality, under the command of an officer duly commissioned by the government of the State . . . and manned by a crew which is under regular armed services discipline.” The latest Navy Commander’s Handbook for Naval Operations asserts that government-owned USVs engaging in noncommercial service are “sovereign immune.” Given this ambiguity, some legal experts conclude that USVs will never be considered warships. This leaves the door open for coastal nations to argue that they should have a right to seize and inspect unmanned vessels.
Safety of Navigation
Perhaps the most significant matter relative to gray zone warfare and USVs is the duty or ability of other mariners to determine if a USV is operating safely or presents a hazard to navigation. Even though safety of navigation featured prominently in each aforementioned USV seizure, there is little legal discussion regarding USVs on the question of safety and hazards to navigation that USVs might present to other vessels.
Professional mariners have been conditioned not to trust automation. The Navy has done its fair share to perpetuate this distrust by building a strong safety record on a conservative approach to redundancies in human and equipment controls. As a result, it has been one of the longest holdouts across the maritime industries in fully embracing the use of technologies like electronic charts, automated radar-plotting aids, and autopilot navigational modes. While there is merit in balancing risk management with innovation, this mindset has inadvertently created an opportunity for U.S. adversaries to plausibly claim an unmanned system is operating erratically or malfunctioning in some way.
The impact of USV seizures on customary law will be borne out over time depending on whether they continue and whether a measured, benign U.S. response becomes the norm. This has also been the case when adversaries have attacked U.S. unmanned aerial vehicles (UAVs), such as Iran’s destruction of a U.S. RQ-4 Global Hawk in 2019. The U.S. argument that USVs have the same legal and sovereign immunity status as manned vessels is undermined when the U.S. response to a USV seizure is far less aggressive than if it were a manned vessel. There has not been enough legal debate and clarity surrounding the defense of USVs, whether that defense is undertaken by the USV autonomously, remotely, or by an altogether different entity like a notional overwatch or quick-reaction force.
Although the U.S. Navy ultimately retrieved all the USVs that were seized by Chinese and Iranian forces, the incidents highlight the need to protect these systems because in the future they are likely to be even more expensive, outfitted with sensitive or classified equipment, or even carrying weapon systems. Anything seized would then be vulnerable to exploitation or reverse-engineering by adversaries.
Beyond the obvious reasons for protecting USVs, the second-order effects of failing to protect them are also worth considering. For instance, the United States may risk giving up tactics, techniques, and procedures to adversaries if a USV is captured. Or, if the United States loses a USV, violent extremist organizations could learn how to make their own delivery devices for weapons of mass destruction.
Recommended Policy and Posture Changes
The Navy must incorporate USV protection into future concepts of operation and procurement programs. The possible solutions for countering adversaries that use gray zone actions against Navy USVs are not likely to be palatable because they will probably constrain the tactical employment of USVs and hamstring incorporating unmanned technologies into naval operations. Despite this, the Navy should explore an array of possible policies designed to protect USVs in tactical environments. These policies could include:
- Manned warships might provide overwatch for a certain number of USVs. This may also force a debate over the span of control for each class of warship. For example, how many USVs can a destroyer provide overwatch for? Or, how close must a coastal patrol boat be to support protecting a USV?
- Maintain regional quick-reaction force fixed with an appropriate tether to respond quickly should a USV suffer engineering or navigational casualties or be interfered with in some way by another vessel.
- Embark force-protection personnel on board larger USVs. This may drive modifications to ship designs for USVs to support living spaces for embarked security personnel. In a recent panel discussion, Chief of Naval Operations Admiral Michael Gilday hinted that the Navy “might make larger unmanned vessels ‘initially minimally manned’ to provide protection.”
- Selectively arm USVs so that they can protect themselves. Experts have even suggested that some type of explosive self-destruct mechanism might be employed on USVs for deterrence to prevent classified or sensitive material from falling into the hands of enemies.
USV employment is at a critical juncture between innovation and gray zone warfare, and the Navy must be prepared to incorporate USV protection into concepts of operation and procurement programs or risk having their tactical use constrained to the point of irrelevance. Manning “unmanned” vessels sounds absurd, would add costs to programs, and would erode many of the efficiencies sought by unmanned systems, whether the manpower is physically on the vessel or is tethered nearby in an overwatch or quick-reaction force role. However, without properly incorporating human protection of USVs, the unique gray-zone legal ambiguity in which USVs operate is likely to constrain USV tactical employment to the point that USV use will diminish rather than expand.