Executive Summary

BLUF: The Corsair has crossed the decisive threshold from experimental unmanned vessel to operational combat system.
On 13 July 2026, USCENTCOM documented three Corsair autonomous surface vessels striking Bandar Abbas Naval Base—the first acknowledged American employment of sea drones in combat.
The platform combines 1,000+ nautical-mile range, 35+ knot speed, and a 1,000-pound modular payload in a 24-foot hull.
The claim that approximately 300 boats were produced or delivered within six months remains unverified by accessible Pentagon procurement records or a traceable Saronic primary-source release.
Likewise, the reported $392 million December 2025 contract cannot presently be validated through an exact, publicly accessible primary-source award notice and is therefore excluded as established fact.
Uncrewed vessels reduce the number of exposed operators; they do not guarantee zero casualties, because strikes against ports, ships, infrastructure, communications nodes, or air-defence positions can kill military personnel and civilians.
The next evolution will be heterogeneous maritime swarms combining surface, subsurface, aerial, cyber-electromagnetic, sensing, logistics, deception, rescue, and strike functions.
By 2031, decisive advantage will depend less on the individual boat than on software, resilient communications, distributed command, automated mission allocation, industrial replacement rates, and counter-swarm economics.
The central contest will become a recursive adaptation cycle: swarm → detection → electronic disruption → autonomy hardening → deception → kinetic interception → mass regeneration.

---

300 Sea Drones and the New Economics of Naval Power

Three unmanned boats entering an enemy naval base may appear marginal beside an aircraft carrier, a submarine or a missile salvo. It is not. On 13 July 2026, US Central Command reported that three Saronic Corsair autonomous surface vessels struck Bandar Abbas Naval Base, marking the first acknowledged American combat use of sea drones. The operation exposed a structural shift in maritime power: states can now place sensors, explosives and autonomous software at risk without placing sailors inside the attacking platform. The promise is lower friendly exposure and faster industrial replacement. The danger is a maritime battlespace crowded with expendable machines, compressed decision times and attacks whose sponsors may remain uncertain. The strategic issue is therefore larger than one Iranian port. It concerns who will control the emerging maritime combat cloud, who can defend ports against it, and who will bear the economic and legal costs when autonomous coercion spreads.

The Operational Threshold

The Bandar Abbas episode established an operational fact that years of demonstrations had not: a US autonomous surface vessel designed for production at scale had moved from experimentation into combat. Three Corsair unmanned surface vessels struck the port at Bandar Abbas Naval Base on 13 July 2026, according to Three Corsair Unmanned Surface Vessels Strike Bandar Abbas Naval Base – US Central Command – July 2026.

Corsair is a 24-foot autonomous surface vessel with a published range exceeding 1,000 nautical miles, a maximum speed above 35 knots and a payload capacity of 1,000 pounds, according to Vessels: Corsair – Saronic Technologies – July 2026. Those dimensions matter because they combine strategic mobility with a payload large enough for surveillance equipment, communications relays, electronic-warfare packages, logistics cargo or an explosive charge. The same hull can therefore become a patrol asset, decoy, sensor node or one-way weapon according to its payload and software.

The widely circulated assertion that the United States produced or delivered 300 Corsair boats in six months should not yet be treated as an independently verified procurement fact. Nor does an advertised $392 million Navy contract reveal a reliable unit price: without an exact public breakdown, the figure may include platforms, software, integration, testing, support equipment and services. The crucial verified development is nevertheless substantial. A long-range, modular autonomous boat has entered combat, and the US Navy is constructing the command, industrial and doctrinal architecture needed to employ such systems in numbers.

From Boat to Combat Cloud

The decisive innovation is not the individual vessel but the network connecting many vessels to aircraft, satellites, crewed ships, coastal sensors and command centres. A remotely operated boat controlled continuously by one person removes the crew from danger but does not fundamentally change fleet economics. Personnel requirements still grow almost in parallel with the number of platforms. A distributed combat cloud changes that ratio. Human commanders establish objectives, geographic limits, rules of engagement and abort criteria; software performs navigation, collision avoidance, formation management, sensor allocation and routine task reassignment.

Task Force 59, created by the US Fifth Fleet in September 2021, was established specifically to integrate unmanned systems and artificial intelligence into Middle Eastern maritime operations. Task Group 59.1 followed in January 2024 to move the concept from experimentation toward routine manned-unmanned deployment. The Navy has since tested aerial drones operating from Coast Guard and naval vessels, remote control across long distances, unmanned launch and recovery, and heterogeneous systems sharing surveillance data.

The likely sequence through 2031 is clear:

Individual remote-controlled vessel → networked autonomous patrol group → crewed-uncrewed task unit → mixed surface-air-underwater formation → distributed maritime combat cloud.

In the mature model, an aerial drone detects a contact and acts as a communications relay; a surface vessel approaches for classification; an underwater vehicle monitors acoustic activity; another unmanned craft carries a decoy or jammer; a crewed ship remains outside the most dangerous zone while retaining command authority. This structure distributes sensing and risk without eliminating human responsibility.

The US Department of Defense has reinforced that direction through its Joint All-Domain Command and Control strategy and the Replicator initiative, which treats autonomy software and resilient networking as capabilities in their own right. Yet the institutional foundations remain incomplete. On 15 June 2026, the US Government Accountability Office concluded that the Navy still faced leadership and organisational challenges in rapidly fielding robotic autonomous systems—Robotic Autonomous Systems: Navy Needs to Address Leadership and Organizational Challenges to Meet Urgent Needs – US Government Accountability Office – June 2026.

GAO had previously examined a Navy plan involving 21 robotic autonomous maritime systems and approximately $4.3 billion over five years, warning that operating, sustainment and digital-infrastructure costs were not fully reflected—Navy Shipbuilding: A Generational Imperative for Systemic Change – US Government Accountability Office – March 2025. Hull production alone does not create combat power. Every operational fleet also requires secure radios, satellite capacity, payloads, batteries, engines, technicians, software updates, launch systems, test facilities and trained controllers.

The Defensive Counter-Revolution

The success of sea drones will accelerate investment in counter-uncrewed systems. Harbour defence will become a maritime equivalent of integrated air defence: several sensor and weapon layers connected through one operational picture. Outer zones will use coastal radar, satellites, airborne electro-optical systems, patrol craft and passive acoustic sensors. Intermediate zones will rely on electronic warfare, remotely controlled barriers and interceptor drones. Terminal zones will combine rapid-fire guns, high-energy lasers, high-power microwave systems and physical hardening around fuel storage, submarines, ammunition depots, cranes and command centres.

The first defensive contest concerns detection. Small vessels operate close to the waterline, where waves, weather, coastal clutter and commercial traffic degrade radar performance. The defender must distinguish an armed autonomous boat from fishing vessels, rescue craft, pilot boats, debris and authorised port robots. Machine vision can accelerate classification, but it introduces the risk of false positives and adversarial deception. Attackers can alter hull profiles, lighting, thermal signatures, wake patterns and approach routes. Defenders will answer by combining radar, optical, infrared and acoustic data rather than trusting a single classifier.

Electronic warfare will remain the cheapest first response, but it will not remain a universal solution. A craft dependent on a continuous radio or satellite link may stop or lose control when jammed. A more advanced platform can continue using inertial navigation, radar, stored maps, coastline recognition and machine vision. The contest will therefore move from simple communications denial to navigation deception, data corruption and uncertainty management. Jamming may not destroy the vessel; it may instead break swarm coordination, delay target recognition and isolate individual attackers for kinetic interception.

Directed energy offers the possibility of reversing the attacker’s cost advantage. Lasers can damage cameras, navigation sensors or exposed components at the speed of light and at a marginal cost far below that of a missile. High-power microwave systems may disrupt the electronics of several closely spaced drones. Yet neither technology provides an unlimited shield. Lasers depend on atmospheric conditions, line of sight, electrical generation and cooling. Microwave systems face shielding, electromagnetic-compatibility and reliability constraints. By 2031, the strongest defensive architecture will therefore remain mixed: electronic warfare to disrupt; interceptor drones to investigate or divert; lasers and microwaves to reduce ammunition expenditure; rapid-fire guns to destroy surviving targets; and barriers to protect the final perimeter.

Europe is developing the same cross-domain logic. The European Defence Agency has supported 15 unmanned maritime research projects worth more than €50 million. Its SABUVIS II programme, involving Poland, Germany, Portugal and Slovenia, received €3.7 million and tested coordinated underwater vehicles and mixed formations combining autonomous underwater and surface systems. China’s official defence publications have likewise discussed unmanned boats performing reconnaissance, patrol, interception and coordinated swarm missions. The technology is therefore not an American monopoly but the beginning of a multinational offence–defence cycle.

The Legal Accountability Gap

“No pilot aboard” does not mean “no human responsibility.” Autonomous maritime warfare redistributes human involvement rather than eliminating it. Commanders define missions; intelligence personnel identify targets; developers construct decision systems; technicians maintain sensors; operators supervise execution; political authorities authorise force. If an autonomous craft strikes the wrong vessel, enters a civilian harbour or continues after political leaders seek to cancel the mission, responsibility cannot be transferred to an algorithm.

The central requirement is reversibility: authorities must be able to suspend, redirect or terminate an operation after launch. That becomes harder when communications are denied and vessels continue under pre-authorised rules. Effective governance therefore requires authenticated abort commands, mission-expiry timers, geographic restrictions, clearly defined escalation thresholds and tamper-resistant logs recording orders, sensor inputs, software versions, confidence levels and operator interventions.

The regulatory environment is beginning to move. On 22 May 2026, the International Maritime Organization adopted the first global non-mandatory Maritime Autonomous Surface Ships Code. It entered into effect on 1 July 2026 for cargo ships covered by SOLAS Chapter I—IMO Adopts First Global Code for Autonomous Ships – International Maritime Organization – May 2026. The code addresses operational modes, risk assessment, connectivity, remote operations, navigation, security, search and rescue and the human element. It does not regulate naval combat, but it establishes a global expectation that autonomy must remain demonstrably safe, controlled and auditable.

The IMO roadmap provides for adoption of a mandatory code by 1 July 2030, with intended entry into force on 1 January 2032—Maritime Safety Committee, 109th Session – International Maritime Organization – December 2024. Military adoption will therefore advance faster than binding civilian regulation, leaving a critical five-year period in which operational practice may outpace legal clarity.

Attribution and the Proxy Advantage

Attribution will become one of the weakest links in maritime deterrence. Commercial engines, cameras, satellite terminals, processors and navigation units are available across international markets. A recovered component may identify a supplier but not the actor that launched, programmed or financed the vessel. A state can transfer a platform to a proxy while retaining intelligence support, mission planning or remote control. An adversary can capture a system and redeploy it under a false signature.

Reliable attribution will require four distinct levels of proof. Technical attribution identifies the platform, software and payload. Operational attribution reconstructs its launch point, route and communications links. Organisational attribution identifies the military unit, intelligence service, contractor or proxy involved. Legal attribution determines whether the operation can be assigned to a state or individual under international law.

This distinction matters because autonomous systems compress tactical events into minutes while high-confidence attribution may require days. A government that retaliates immediately risks attacking the wrong actor. A government that waits may invite further attacks and weaken deterrence. The most destabilising proliferation pathway is therefore not necessarily a major navy operating recognisable platforms, but a proxy using commercially assembled vessels with sufficient autonomy to threaten shipping, offshore energy infrastructure or a secondary port.

The Insurance and Port-Security Shock

Marine insurance will translate autonomous risk into economic cost before international law fully resolves it. Underwriters must decide whether a collision or attack resulted from defective software, negligent remote supervision, cyber intrusion, satellite disruption, spoofed navigation or deliberate military action. A common software defect could affect dozens of vessels simultaneously, converting an isolated accident into a correlated portfolio loss.

By 2031, insurance coverage is likely to depend on certified software, redundant navigation, cyber controls, operator qualifications, incident logging, geofencing and clearly identified legal responsibility. Ports may require autonomous vessels to present authenticated digital identities and proof of liability coverage before entry. Vessels unable to demonstrate who controls them may be denied access or confined to designated corridors.

Port security will also become more expensive. The International Ship and Port Facility Security Code was designed around human-controlled ships, cargo, access control and terrorist threats. Autonomous systems introduce crewless penetration, underwater approaches, remote payload activation and cyber-physical attacks against port-management systems. High-value facilities will need controlled autonomous-vessel zones, compulsory telemetry, machine-readable routing instructions, underwater surveillance and remotely activated barriers.

This will divide the maritime economy. Major ports will finance sophisticated protection and retain insurer confidence. Smaller ports may face higher premiums, operational restrictions or traffic diversion. The attacker does not need to destroy an entire terminal; it may be sufficient to create doubt about safe access.

The New Cost of Coercion

Maritime coercion once required submarines, mines, aircraft, anti-ship missiles or major surface forces. Autonomous vessels lower the entry cost. A small number of credible attacks can delay port calls, force rerouting, raise insurance premiums and increase fuel, inventory and financing costs across entire supply chains. The economic effect can greatly exceed the replacement value of the attacking boats.

This produces a structural imbalance. Defenders must spend continuously on radar, acoustic arrays, patrols, cyber defence, barriers, interceptors and trained personnel. Attackers can choose the timing, route and concentration of a raid. The defender protects every day; the attacker needs only one successful penetration—or even the credible threat of one—to disrupt confidence.

The Corsair therefore represents more than a new naval weapon. It signals a change in the economics of presence, attrition and coercion. A 24-foot craft travelling more than 1,000 nautical miles, carrying 1,000 pounds and exceeding 35 knots can force an adversary to protect infrastructure worth billions. Hundreds of such platforms, should large-scale production claims eventually be independently confirmed, would not replace destroyers or submarines. They would create an expendable outer fleet that scouts, deceives, relays, supplies and attacks ahead of them.

The winning navy of 2031 will not necessarily possess the greatest number of autonomous boats. It will possess the strongest combination of secure software, modular payloads, resilient communications, industrial replacement capacity, human command and layered defence. The central strategic danger is equally clear: as the cost of placing machines in danger falls, the political threshold for placing other people, ports and global trade at risk may fall with it.

---

Navigational Index

Pillar I — From Remote-Controlled Boat to Distributed Maritime Combat Cloud

Platform verification, operational significance, modular payload architecture, machine-speed coordination, command-and-control evolution, and the transition from individual vessels to cross-domain swarms.

Pillar II — The Five-Year Offence–Defence Competition

Counter-uncrewed systems, electronic warfare, harbour defence, machine vision, acoustic surveillance, directed energy, interceptor drones, cyber compromise, autonomous navigation under communications denial, and industrial attrition.

Pillar III — Strategic, Legal and Geoeconomic Consequences

Escalation control, human accountability, attribution, insurance, port security, naval force structure, industrial mobilisation, adversary replication, proliferation, and the changing cost of maritime coercion.

Master Abstract

The 13 July 2026 Bandar Abbas operation is strategically important not because three relatively small boats damaged a naval installation, but because it demonstrated the operational convergence of autonomous navigation, long-range deployment, modular payload integration, persistent surveillance and expendable precision attack. An official US Central Command video states that three Corsair unmanned surface vessels struck Bandar Abbas Naval Base and identifies the episode as the first American combat employment of sea drones—USCENTCOM Video Gallery – United States Central Command – July 2026 — verified primary-source record. The attached narrative supplied for this assessment adds claims concerning a prior rescue mission, a $392 million acquisition, approximately 300 vessels, a strike against port infrastructure and damage to an Iranian submarine; those assertions require disaggregation because they do not possess equal evidentiary strength. The combat strike itself is officially documented, while the quantity, contract value, full battle-damage assessment, autonomous-control level and earlier rescue account cannot all be independently confirmed through the permitted primary-source hierarchy. What is verifiable is the underlying platform envelope: Saronic identifies Corsair as a 24-foot autonomous surface vessel with a range exceeding 1,000 nautical miles, speed exceeding 35 knots, and payload capacity of 1,000 pounds—Vessels: Corsair – Saronic Technologies – accessed July 2026 — verified manufacturer specification. Those specifications create a militarily significant combination of reach, mobility and payload flexibility. A vessel of this type can function as a sensor picket, communications relay, electronic-warfare carrier, decoy, logistics shuttle, casualty-recovery platform, mine-countermeasure node or one-way attack system. The strategic transformation therefore lies in mission fungibility: a common hull can be reassigned through payload substitution and software configuration, enabling commanders to change force composition without redesigning the entire vessel. This architecture compresses procurement, training and deployment cycles while complicating adversary intelligence, because identifying the hull does not necessarily reveal the mission.

The user’s proposition that an unmanned boat means “no human pilot inside, no casualties, only damage to adversaries” captures the political attraction of remote warfare but not its operational or legal reality. Removing a crew from the attacking platform reduces friendly exposure, eliminates the need to recover onboard personnel and makes commanders more willing to risk the vehicle; it does not remove human beings from the kill chain, the battlespace or the consequences. Remote operators, mission commanders, intelligence analysts, software engineers, communications personnel and authorising officers remain integral to the system. Target areas may contain sailors, guards, maintenance workers, emergency personnel, contractors or civilians. Autonomous vessels may also malfunction, collide with commercial shipping, lose navigation integrity, misclassify objects or be captured and exploited. The decisive distinction is therefore not between “human” and “machine” warfare, but between human presence aboard the platform and human responsibility for the mission. International maritime regulation is already moving toward explicit requirements for operational modes, safety equivalence, remote-control arrangements, cyber resilience and accountability. The International Maritime Organization adopted the non-mandatory Maritime Autonomous Surface Ships Code in May 2026, effective from 1 July 2026, to provide a goal-based safety, security and environmental framework for remotely controlled and autonomous commercial ships—IMO Adopts First Global Code for Autonomous Ships – International Maritime Organization – May 2026 — verified institutional source. The Code does not govern combat employment as such, but its treatment of operating modes, system assurance, communications, remote-control centres and the human element anticipates the regulatory problems military fleets will face. The IMO roadmap also points toward a mandatory code targeted for adoption by 1 July 2030 and entry into force in 2032—Maritime Safety Committee, 110th Session – International Maritime Organization – June 2025 — verified regulatory roadmap. Military adoption will move faster than civilian law, creating a five-year interval in which operational practice may outrun shared norms governing collision responsibility, distress assistance, identification, communications failure and control transfer.

The most probable five-year evolution is not a linear progression toward a single “better boat,” but a transition toward distributed maritime machine forces whose combat power emerges from networking, mission allocation and replacement capacity. The US Navy has already treated Task Force 59 as an experimentation environment for integrating unmanned systems into fleet operations, while the service’s broader surface strategy identifies unmanned platforms as tools for extending sensing, sea denial and combat reach—Surface Warfare: The Competitive Edge 2.0 – Commander, Naval Surface Forces – January 2025 — verified US Navy strategy document. Yet institutional scaling remains a vulnerability. A June 2026 Government Accountability Office assessment reported that the Navy had consolidated robotic and autonomous-system acquisition under a portfolio executive, indicating an effort to correct fragmented governance but also confirming that acquisition, life-cycle management and integration remain unresolved institutional tasks—Robotic Autonomous Systems: Navy Actions – US Government Accountability Office – June 2026 — verified oversight report. European programmes are following a parallel trajectory: the European Defence Agency reported mixed-swarm trials involving coordinated movement, data exchange, formation control and adaptive mission execution across multinational systems—EDA Project Develops Technology for Underwater Drones to Move in Swarms – European Defence Agency – February 2026 — verified EU defence source. Chinese official reporting likewise describes unmanned boats, autonomous underwater vehicles and unmanned mine-laying systems capable of concealed deployment, autonomous detection, networked swarming and coordinated attack—Victory Day Parade Equipment Overview – Ministry of National Defense of the People’s Republic of China – September 2025 — verified Chinese government source. These independently observable trajectories support a high-confidence assessment: by 2031, maritime autonomy will no longer be a specialist adjunct to conventional fleets. It will form an attritable outer layer that scouts, deceives, jams, supplies, mines, intercepts and attacks ahead of crewed vessels, forcing every major navy and port authority to develop permanent counter-autonomy architectures.

From Remote-Controlled Boat to Distributed Maritime Combat Cloud

The transition from a remotely operated boat to a distributed maritime combat cloud begins with a strict separation between what is operationally verified, what is technically plausible, and what remains promotional or wartime reporting. US Central Command has officially documented the employment of three Corsair unmanned surface vessels against Bandar Abbas Naval Base on 13 July 2026, describing the action as the first American combat use of sea drones. This establishes the operational event, but it does not independently establish every reported detail concerning target identity, battle damage, control mode, launch location, route, electronic-warfare conditions, or the alleged size of the deployed fleet—Video Gallery: Three Corsair Unmanned Surface Vessels Strike Bandar Abbas Naval Base – United States Central Command – July 2026 — verified official operational record. The supporting narrative provided for this study claims approximately 300 boats, a $392 million procurement, an earlier combat search-and-rescue mission, and damage to Iranian naval infrastructure and a submarine; those propositions must be treated as separate hypotheses rather than merged into one confirmed account. The manufacturer publicly states that Corsair is a 24-foot autonomous surface vessel capable of carrying a 1,000-pound payload, travelling more than 1,000 nautical miles, and exceeding 35 knots—Vessels: Corsair – Saronic Technologies – accessed July 2026 — verified manufacturer specifications. Saronic’s website also advertises a $392 million Navy contract, but publicly accessible primary material reviewed for this section does not provide sufficient detail to calculate vessel quantity, unit cost, payload configuration, delivery schedule, or the proportion allocated to Corsair rather than associated software, integration, support, test equipment or other platforms—Redefining Maritime Superiority – Saronic Technologies – accessed July 2026 — verified corporate announcement listing. Consequently, the strategically sound baseline is not “the United States unquestionably built 300 attack boats in six months,” but rather that an operationally employed, long-range, high-payload autonomous boat exists, has entered combat, and belongs to an industrial programme explicitly oriented toward production at scale.

That distinction matters because the military value of Corsair does not principally reside in the hull. A 24-foot, fast, relatively low-observable vessel is useful, but the transformative element is a modular architecture in which propulsion, navigation, perception, communications, mission software and payload interfaces can be recombined across different operational packages. The same basic platform could theoretically conduct maritime-domain awareness, electronic surveillance, communications relay, reconnaissance, deception, logistics, casualty recovery, harbour penetration, mine countermeasures or one-way attack, although only functions supported by verified public evidence should be treated as fielded capabilities. The manufacturer’s specification of a 1,000-pound payload allowance creates sufficient physical margin for alternative sensor, fuel, communications, countermeasure or effect packages, but payload capacity alone does not prove that a specific weapon, radar, jammer, torpedo, mine or rescue module has been certified—Vessels: Corsair – Saronic Technologies – accessed July 2026 — verified platform data. The operational logic resembles a software-defined naval truck rather than a miniature conventional warship: common hulls generate scale, while payload and code determine mission identity. This architecture can reduce training diversity, simplify spares, accelerate reconfiguration and obscure the adversary’s understanding of each contact. An opposing commander observing ten externally similar vessels may not know whether they carry cameras, relays, decoys, electronic-warfare equipment, explosive payloads or reserve fuel for extended surveillance. That uncertainty imposes a defensive tax, because every contact may require classification, tracking and potentially interception. However, modularity also creates engineering liabilities. Centre-of-gravity changes, power requirements, electromagnetic interference, cooling, saltwater exposure, software certification, communications bandwidth and signature management vary across payloads. A platform cannot become genuinely modular merely by possessing deck space; it requires tested mechanical, electrical, data and safety standards that permit rapid payload exchange without destabilising navigation, degrading endurance or creating cyber vulnerabilities. The decisive five-year metric will therefore be neither hull count nor payload weight in isolation, but the number of mission packages that can be installed, validated and operationally reassigned within hours rather than months.

The command evolution from one operator controlling one boat toward one mission commander supervising tens or hundreds of agents represents the central technical and organisational discontinuity. Legacy unmanned systems often reproduce manned-force staffing in remote form: one platform, one control station, several specialists and continuous communications. Such a model may remove personnel from the hull but does not generate scalable autonomy because manpower requirements increase almost linearly with platform numbers. The future combat cloud instead requires supervisory control, under which humans define mission objectives, geographic boundaries, rules, priorities, abort criteria and authorisation thresholds while software handles navigation, collision avoidance, sensor management, formation keeping, task distribution and routine contingency responses. The Department of the Navy Unmanned Campaign Framework explicitly identifies command-and-control structure, intended employment, degrees of autonomy and continued human responsibility under the law of armed conflict as core design issues—Department of the Navy Unmanned Campaign Framework – United States Navy – March 2021 — verified official framework. NAVSEA has separately depicted a transition away from incompatible one-to-one control solutions toward a common control system integrated into shore facilities and shipboard combat systems—PMS 406 Unmanned Maritime Systems – Naval Sea Systems Command – April 2022 — verified official programme presentation. This is the foundational infrastructure of a maritime combat cloud: operators should not need to learn a proprietary interface for every vendor or platform. They require a common operational picture, common messaging standards, common identity management, common mission planning and a controlled method for transferring authority between shore, ship, aircraft and forward command nodes. Without that layer, a large unmanned fleet becomes a collection of disconnected remote-control systems vulnerable to bandwidth saturation, operator overload and incompatible data formats. With it, the platforms become interchangeable nodes inside a distributed force.

Verified platform-to-cloud progression

Development layer	Remotely controlled boat	Networked autonomous vessel	Distributed maritime combat cloud
Human role	Continuous steering and sensor supervision	Mission-level supervision with intervention	Command intent, force allocation and engagement authorisation
Control relationship	One operator or crew per platform	One team supervising several platforms	Distributed teams supervising heterogeneous formations
Communications	Persistent high-bandwidth control link	Intermittent links with local autonomy	Multi-path, multi-band, cross-domain resilient networking
Mission logic	Pre-planned route or direct control	Dynamic navigation and limited task adaptation	Machine-assisted mission allocation and collaborative behaviour
Sensor use	Platform-local observation	Shared tracks among selected nodes	Fused maritime, air, space, cyber and acoustic picture
Payload employment	Single-purpose configuration	Modular mission package	Cross-domain sequencing of sensing, deception, logistics and effects
Failure response	Platform stops or awaits operator	Local fallback and return behaviour	Swarm reorganisation, task reassignment and network healing
Operational value	Remote presence	Attritable capability	Distributed sensing, decision and action architecture

Task Force 59 is best understood as an operational laboratory for this transition rather than merely a unit that owns unmanned boats. Established in September 2021, it was explicitly designed to integrate unmanned systems and artificial intelligence into maritime operations in the US Fifth Fleet area—US Fifth Fleet Launches New Task Force to Integrate Unmanned Systems – US Naval Forces Central Command – September 2021 — verified official establishment notice. In January 2024, the Navy activated Task Group 59.1 to emphasise operational deployment of unmanned systems teamed with manned operators rather than episodic experimentation—Task Force 59 Launches New Unmanned Task Group 59.1 – United States Navy – January 2024 — verified official Navy release. The command’s official record indicates that it has operated aerial, surface and underwater systems and has experimented with resilient communications, remote launch, shipboard integration and machine-assisted maritime monitoring. Its use of the Flexrotor unmanned aircraft was specifically associated with relaying imagery and data into command centres ashore and at sea, demonstrating that the aircraft could serve simultaneously as an intelligence platform and a communications node—Task Force 59 Launches Aerial Drone from Coast Guard Ship in Middle East – United States Navy – December 2022 — verified Navy operational release. By August 2025, the command reported an autonomous launch and recovery of a Group II aerial vehicle from an unmanned surface vessel and a remote launch of a loitering strike munition from an unmanned system, both relevant precursors to heterogeneous swarm operations—Task Force 59 Holds Change of Command Ceremony – US Naval Forces Central Command – August 2025 — verified command record. These events show that the combat cloud is emerging through functional coupling: an unmanned boat can carry, launch, recover, relay for or cue another unmanned system, thereby becoming a mobile edge-computing and mission-support node rather than an isolated vessel.

The transition to machine-speed coordination will occur incrementally because fully centralised and fully decentralised architectures each contain critical weaknesses. A central command node can optimise the entire formation using a comprehensive operational picture, but it creates a high-value point of failure and demands substantial bandwidth. A decentralised swarm can continue functioning when communications degrade, but distributed decisions may diverge, produce duplicate tasking, interfere with friendly forces or generate emergent behaviour that commanders cannot readily predict. The most credible architecture is therefore hierarchical and federated. At the top level, human commanders assign operational objectives and legal constraints. A mission-management layer decomposes those objectives into surveillance sectors, routes, timing windows, electronic-support tasks, decoy behaviours, logistics missions or authorised effects. Local swarm leaders or rotating coordinators then distribute tasks among nearby nodes based on fuel, payload, sensor health, communications quality and threat exposure. Each vessel retains enough autonomy to avoid collisions, navigate around hazards, preserve itself when appropriate and return or loiter when connectivity fails. The Joint All-Domain Command and Control concept seeks to connect sensors and effectors across domains through data integration and artificial intelligence, providing the wider institutional logic within which maritime autonomous nodes would operate—Summary of the Joint All-Domain Command and Control Strategy – United States Department of Defense – March 2022 — verified strategy summary. The Department’s Replicator 1.2 announcement also stated that selected capabilities would include air and maritime systems together with software enablers intended to improve autonomy and resilience—Deputy Secretary of Defense Announces Additional Replicator All-Domain Attritable Autonomous Capabilities – United States Department of Defense – November 2024 — verified official release. This implies that software interoperability, resilient networking and autonomy services are becoming acquisition objects in their own right, rather than incidental components attached to individual vehicles.

The decisive technical constraint is no longer maximum range or peak speed but the ability to preserve mission coherence inside an adversarial electromagnetic and cyber environment. A boat advertised with 1,000 nautical miles of range cannot exploit that distance if it depends on a constant, high-bandwidth connection that an adversary can jam, geolocate, spoof or penetrate. Combat-cloud resilience therefore requires multiple communications pathways, graceful degradation and mission continuity under partial isolation. A surface vessel may communicate through direct radio, satellite links, airborne relays, other vessels or temporary store-and-forward networks. When one pathway disappears, the system must determine whether to continue, pause, return, transfer control or execute a pre-authorised contingency. This requires local state estimation, trusted timekeeping, inertial navigation, terrain or coastline matching, radar and electro-optical navigation, authenticated software, encrypted mission data and protection against adversarial manipulation. The Department of Defense’s satellite command-and-control planning calls for the ability to dynamically re-plan and reallocate satellite resources across military, commercial and non-satellite networks, reflecting the broader requirement for communications agility—Enterprise Satellite Communications Management and Control Implementation Plan – United States Department of Defense Chief Information Officer – January 2023 — verified official implementation plan. Cybersecurity must extend beyond encryption to supply-chain assurance, secure boot, signed updates, hardware identity, anti-tamper protection, intrusion detection and compartmented access to payload controls. The Department of Defense Directive 3000.09 requires autonomous and semi-autonomous weapon systems to incorporate verification, validation, system safety, cybersecurity, anti-tamper measures, operator training and appropriate levels of human judgment over force—Autonomy in Weapon Systems, DoD Directive 3000.09 – United States Department of Defense – January 2023 — verified governing directive. A combat cloud that cannot prove the identity, software state, authority and sensor provenance of each participating node risks becoming a mechanism through which an adversary injects false tracks, redirects platforms or triggers unintended interactions.

European programmes confirm that the movement from individual unmanned vehicles to cooperative formations is not exclusively American. In February 2026, the European Defence Agency announced completion of the second phase of its SABUVIS II project, involving Poland, Germany, Portugal and Slovenia, with a reported budget of €3.7 million. The project focused on coordinated autonomous underwater vehicles, scalable lower-cost swarms, operation in shallow or cluttered littoral environments and mixed formations combining underwater vehicles with autonomous surface craft—EDA Project Develops Technology for Underwater Drones to Move in Swarms – European Defence Agency – February 2026 — verified European Union defence source. The importance of this work lies less in any single vehicle than in formation control, collaborative navigation, underwater communications, cooperative sensing and adaptive mission behaviour. Underwater agents cannot rely on continuous high-bandwidth radio or satellite links; acoustic communications are slower, less reliable and more detectable. Consequently, underwater swarms must possess greater local autonomy and tolerate longer periods of information isolation. The EDA has also identified network-enabled cooperation, persistent autonomy and underwater communication as central challenges in unmanned maritime mine-countermeasure programmes—Autonomy in Defence: Systems, Weapons and Decision-Making – European Defence Agency – accessed July 2026 — verified EDA programme overview. European experimentation therefore supports a broader inference: the mature combat cloud will not function as a single homogeneous swarm in which every platform behaves identically. It will consist of heterogeneous clusters governed by different communications tempos. Aerial nodes will provide rapid relay and wide-area surveillance; surface nodes will offer endurance, payload and gateway functions; underwater nodes will supply concealment, acoustic intelligence, mine warfare and anti-submarine coverage. The command architecture must reconcile these different speeds of sensing, movement and communication without overwhelming operators or producing contradictory tracks.

Chinese official military discourse has for years treated unmanned boats as components of integrated reconnaissance, precision attack and swarm operations. A Chinese Ministry of National Defense publication defines intelligent unmanned boats as remotely controlled or autonomous surface combat platforms and divides them into attack, reconnaissance-patrol and other mission categories—Future Application Prospects of Intelligent Unmanned Boats – Ministry of National Defense of the People’s Republic of China – June 2021 — verified Chinese official source, Chinese language. Another official Chinese defence article describes unmanned surface swarm concepts involving different vessel classes cooperating in autonomous missions—Combat Unmanned Swarms: Preparing to Take Flight – Ministry of National Defense of the People’s Republic of China – April 2020 — verified Chinese official source, Chinese language. Official Chinese reporting has also displayed a missile-armed unmanned boat equipped with electro-optical and radar systems and a four-cell launcher, demonstrating longstanding interest in combining surveillance and precision strike on a small autonomous hull—China’s First Missile-Armed Unmanned Boat Publicly Displayed – Ministry of National Defense of the People’s Republic of China – November 2018 — verified Chinese official source, Chinese language. These sources do not prove current Chinese fleet quantities, combat readiness or specific classified architectures; they do show that Beijing conceptually recognises the same trajectory from remote control to collaborative autonomy. Russian official reporting accessible through the Russian Ministry of Defence media environment places stronger public emphasis on defeating drones through manoeuvre, electronic warfare and tactical adaptation than on disclosing a comparable maritime combat-cloud architecture—On the Water It Is Sometimes More Dangerous Than During an Assault on Land – Russian Ministry of Defence Media Resource – April 2026 — verified Russian official-domain source, Russian language. The asymmetry in public disclosure should not be misread as absence of capability; it instead indicates that open-source confidence concerning Russian maritime-swarm command systems remains substantially lower than confidence concerning declared American, European and Chinese programmes.

Analysis of competing hypotheses: 2026–2031

Hypothesis	Core proposition	Evidence increasing probability	Evidence decreasing probability	Updated probability
H₁: Hull-Mass Dominance	Large quantities of inexpensive boats independently overwhelm maritime defences	Replicator-scale ambitions; combat employment; modular small craft	Bandwidth, maintenance, launch and operator constraints	19%
H₂: Network Primacy	Data architecture and resilient C₂ determine combat value more than raw platform count	JADC2, common-control development, resilient communications, cross-domain integration	Cyber vulnerability and interoperability friction	28%
H₃: Payload-Modularity Revolution	Common hulls with rapidly exchanged payloads transform naval force generation	Corsair payload margin; Navy modular-control approach	Certification, power, weight and integration difficulties	17%
H₄: Crewed–Uncrewed Hybrid Fleet	Autonomous systems amplify rather than replace crewed ships and aircraft	TF 59 manned-unmanned teaming; operational command requirements	Political pressure to minimise crew exposure	25%
H₅: Defensive Adaptation Cancels the Advantage	Harbour barriers, EW, interceptors and automated detection neutralise small USVs	Low-altitude maritime detectability; evolving counter-drone systems	Defender cost, geography and saturation pressure	11%

The Bayesian assessment assigns the highest combined probability to H₂ and H₄, because verified institutional development is concentrating on common control, data exchange, resilient communications and manned-unmanned teaming rather than on unsupervised flotillas operating independently. These probabilities are analytical estimates derived from weighted evidence, not official forecasts. The prior distribution began at 20% for each hypothesis. Verified combat employment raised H₁ and H₃, but the absence of independently confirmed fleet quantities prevented a larger upward update for mass dominance. Navy and Department of Defense documentation on common-control software, JADC2, Replicator software enablers and resilient communications raised H₂. TF 59’s operational pattern—integrating unmanned aircraft, surface vessels, Coast Guard ships, littoral combat ships, shore centres and human operators—raised H₄. European mixed-swarm experimentation further supported network primacy and cross-domain integration. H₅ remains lower in the near term because no public primary-source record demonstrates that harbour defences have already neutralised the class as a whole; nevertheless, its probability rises over the five-year period as defenders automate detection and interception. The key judgement is that maritime autonomy will produce advantage through precise distribution, not merely numerical abundance. A hundred disconnected boats may possess less operational value than twenty vessels connected to airborne relays, satellite surveillance, crewed ships, electronic-intelligence systems and a common mission-management layer. Conversely, excessive dependence on a single cloud, satellite constellation or proprietary software service would create systemic fragility. The winning architecture will distribute both physical platforms and decision-support functions, allowing local agents to continue operating when the wider network becomes intermittent while preserving human control over mission boundaries and lethal authorisation.

Five-year capability timeline

Period	Most likely operational development	Principal enabling technology	Dominant failure risk	Assessment confidence
2026–2027	Expansion from combat demonstration to persistent patrol, relay, reconnaissance and selected strike packages	Common control interfaces, modular payload integration, improved production	Proprietary systems remain incompatible	High
2027–2028	One operator supervises multiple surface vessels; aerial relays routinely support maritime formations	Supervisory autonomy, machine-assisted route and task planning	Operator overload shifts from steering to exception management	High–medium
2028–2029	Mixed surface-air-underwater teams conduct coordinated sensing and deception	Distributed track fusion, mesh networking, edge computing	False tracks and cyber compromise contaminate the shared picture	Medium
2029–2030	Formations continue missions through intermittent communications and satellite denial	Multi-modal navigation, local mission reasoning, network healing	Autonomous behaviours diverge under degraded information	Medium
2030–2031	Combat clouds dynamically reassign sensors, relays, decoys, logistics and authorised effectors	Federated mission management, machine-speed deconfliction, auditable autonomy	Escalation and accountability become harder to reconstruct	Medium–low

Industrial scaling will determine whether this architecture exists continuously in theatre or only during demonstrations. The Government Accountability Office reported in June 2026 that recent conflicts had accelerated naval interest in robotic and autonomous systems but that the Navy still faced leadership and organisational challenges in meeting urgent needs—Robotic Autonomous Systems: Navy Needs to Address Leadership and Organizational Challenges to Meet Urgent Needs – US Government Accountability Office – June 2026 — verified federal oversight report. Earlier GAO analysis found that a Navy plan for 21 robotic autonomous maritime systems and approximately $4.3 billion over five years did not fully account for operations, sustainment and necessary digital infrastructure, warning that incomplete cost estimates could distort affordability assessments—Navy Shipbuilding: A Generational Imperative for Systemic Change – US Government Accountability Office – March 2025 — verified federal oversight source. This finding is central because a distributed combat cloud requires far more than hull production. It requires batteries, engines, sensors, satellite airtime, secure radios, test ranges, software integration, cyber certification, launch-and-recovery equipment, maintenance personnel, spares, storage, transport, mission-data preparation and training. A nominally inexpensive vessel may become operationally costly if it demands specialised technicians, contractor support or constant high-bandwidth satellite control. Conversely, a more expensive platform may generate superior lifecycle value if it shares components, can be repaired forward, supports multiple payloads and operates under intermittent communications. The claim of 300 boats in six months, even were it eventually verified, would therefore describe production throughput rather than usable combat power. The relevant denominator is mission-capable systems available at a given place and time, with functioning payloads, trusted software, trained controllers, communications access and an approved operational concept. By 2031, procurement authorities are likely to distinguish more sharply between the purchase price of autonomous hulls and the recurring cost of the combat cloud that makes them useful.

The strategic endpoint is not a navy without people but a navy in which humans are progressively displaced from the most exposed physical positions and concentrated at the levels of intent, judgement, exception management and accountability. DoD Directive 3000.09 requires commanders and operators to retain appropriate levels of human judgment over the use of force, while autonomous systems must be tested, verified and employed within approved doctrine, safety rules and rules of engagement—Autonomy in Weapon Systems, DoD Directive 3000.09 – United States Department of Defense – January 2023 — verified directive. The civilian regulatory environment is also establishing concepts that will indirectly influence military assurance practices. In May 2026, the International Maritime Organization adopted a non-mandatory Maritime Autonomous Surface Ships Code, effective 1 July 2026, covering operational modes, risk assessment, software principles, connectivity, remote operations, navigation, security, search and rescue and the human element for applicable commercial vessels—IMO Adopts First Global Code for Autonomous Ships – International Maritime Organization – May 2026 — verified international regulatory source. The IMO’s roadmap targets adoption of a mandatory code by 1 July 2030, with intended entry into force on 1 January 2032—Maritime Safety Committee, 110th Session – International Maritime Organization – June 2025 — verified regulatory roadmap. Military systems remain governed by different legal regimes, but the same technical questions recur: Who is the master or commander? Which node possesses authority? How is control transferred? What happens after communications loss? How are decisions logged? How is software configuration certified? How does the vessel distinguish a lawful target, protected object, civilian traffic or person in distress? The mature combat cloud will require an evidentiary architecture capable of reconstructing the state, orders, sensor inputs, software version and communications history of each node. Without such auditability, machine-speed coordination may improve tactical tempo while degrading strategic control and post-event accountability.

The five-year outlook therefore supports a high-confidence judgement that individual unmanned boats will evolve into cross-domain force packages, but only a medium-confidence judgement that they will operate as large, highly decentralised lethal swarms by 2031. Persistent maritime sensing, communications relay, decoy activity, logistics, mine countermeasures and supervised multi-vessel control are technically and institutionally closer than unconstrained autonomous combat. The most likely sequence is individual platform → networked patrol group → manned-unmanned task unit → heterogeneous surface-air-underwater formation → federated combat cloud. Each stage increases operational reach while transferring complexity from the hull into software, networks, data governance and command procedures. The attached source correctly identifies production and combat employment as strategically important, but its strongest claims should remain provisional until exact primary-source procurement and battle-damage evidence becomes available. The future fleet’s decisive competitive variables will be mission-package interchangeability, autonomous operation during communications denial, secure identity, shared track quality, software-update speed, forward repair, launch-and-recovery capacity and the ratio between platforms and human supervisors. The greatest conceptual error would be to assume that uncrewed means casualty-free, cheap, autonomous or strategically controllable. Uncrewed hulls reduce onboard personnel exposure; they can simultaneously lower political thresholds for force employment, create new risks to civilians and commercial navigation, widen proliferation and increase the speed at which local incidents become regional crises. The combat cloud will be strategically valuable precisely because it distributes presence, sensing and effects. It will be dangerous for the same reason: once hundreds of adaptive nodes share information and react at machine speed, the principal challenge becomes not making them act, but ensuring that they act only within the commander’s lawful and politically intended boundaries.

Pillar II — The Five-Year Offence–Defence Competition

The offence–defence competition around unmanned maritime systems will be determined by whether attackers can preserve low-signature access, navigational coherence and payload delivery long enough to cross a progressively denser defensive perimeter. The defender’s problem is structurally harder than simply “shooting the boats” because a harbour or naval formation must distinguish hostile unmanned surface vessels from fishing craft, pilot boats, debris, civilian traffic, rescue platforms, decoys and friendly autonomous systems before authorising force. Small craft operate close to the sea surface, where radar performance is degraded by waves, multipath reflections, precipitation, spray, coastline clutter and the curvature-limited horizon. Attackers can exploit commercial hull geometries, reduced radar cross-sections, low thermal emissions, pre-programmed routes, deceptive transponder behaviour and staggered approaches from multiple bearings. Defenders, by contrast, must maintain persistent coverage and avoid catastrophic false positives. The American response already points toward layered integration rather than a single countermeasure. In February 2026, the Naval Surface Warfare Center Dahlgren Division announced delivery of the MK 38 Mod 4, integrating a 30 mm gun with the Aegis Combat System and explicitly strengthening defence against high-speed, manoeuvrable unmanned surface vessels as well as unmanned aircraft—NSWCDD Delivers Next-Generation MK 38 Defense System – Naval Sea Systems Command – February 2026. This development illustrates the likely near-term defensive architecture: existing ship combat systems will absorb new sensors, machine-assisted classification and lower-cost effectors rather than depend exclusively on strategic missiles or manually aimed weapons. The tactical goal will be to create multiple engagement opportunities—detection outside the harbour, classification during approach, non-kinetic disruption before weapons release, interception at intermediate range and kinetic destruction inside the terminal zone—while preserving enough ammunition and electrical power to defeat repeated waves rather than a single spectacular raid.

Harbour defence will evolve into a multi-layered sensor-and-effector mesh in which geometry matters as much as technology. Outer surveillance zones will exploit coastal radar, electro-optical towers, airborne sensors, commercial satellite products, automatic identification data, patrol vessels and seabed or buoy-mounted acoustic arrays. Intermediate zones will use chokepoints, floating barriers, controlled navigation lanes, remotely operated inspection craft, electronic-warfare emitters and interceptor drones. Inner zones will protect fuel farms, ammunition depots, dry docks, command nodes, submarines and critical cranes with physical barriers, rapid-fire guns, short-range missiles, laser dazzlers, high-energy lasers, high-power microwave systems and hardened blast separation. The European Defence Agency identifies counter-unmanned technologies, advanced sensing, intelligence-surveillance-reconnaissance capabilities, autonomous maritime systems, electromagnetic weapons and hypervelocity effectors as converging priorities within its maritime technology portfolio—CapTech Maritime – European Defence Agency – accessed July 2026. The same institution’s unmanned-maritime research programme has launched 15 projects with a combined value exceeding €50 million, demonstrating that European states are investing in the enabling technologies needed both to operate and counter unmanned systems—Unmanned Maritime Systems Research – European Defence Agency – accessed July 2026. The defensive challenge, however, is not solved by adding sensors indiscriminately. Every additional radar, camera, sonar and acoustic node creates data-fusion, latency, cybersecurity and operator-attention burdens. A harbour may possess dozens of sensors yet remain vulnerable if tracks cannot be correlated, ownership cannot be established and warning thresholds are inconsistent. The mature architecture will therefore require a unified track identity that follows a contact from offshore detection through terminal engagement, recording confidence scores, sensor provenance, classification changes and rules-of-engagement status. By 2031, the most capable ports will resemble integrated air-defence networks translated into the maritime littoral: layered, automated, geographically distributed and designed to preserve continuity even after individual sensors, communications links or command posts are lost.

Electronic warfare will become the first scalable defensive layer because it can potentially disrupt several vessels without expending one missile or projectile per target. Yet the popular assumption that jamming automatically disables an unmanned boat is increasingly unreliable. A remotely controlled craft dependent on a continuous radio or satellite link may stop, lose situational awareness or default to a return route when its communications are denied. A more advanced vessel can continue through inertial navigation, visual coastline matching, radar odometry, stored terrain data, local obstacle avoidance and pre-authorised mission logic. Consequently, the offence–defence contest will shift from simple link jamming toward a contest over confidence in navigation and perception. Defenders will attempt to suppress or deceive GNSS, interfere with telemetry, inject false network traffic, exploit control protocols and create deceptive visual or radar features. Attackers will respond with multi-constellation receivers, inertial systems, anti-jam antennas, encrypted timing, sensor fusion and local autonomy. The European Defence Agency identifies GNSS radio-frequency interference, electromagnetic hazards, unmanned systems and counter-unmanned systems as increasingly important operational complications—Electronic Warfare – European Defence Agency – accessed July 2026. Russian official military reporting in April 2026 described Ukrainian small unmanned boats and drones using machine vision, indicating that the attacker’s adaptation path already includes reducing dependence on external navigation and control signals—На воде иногда опаснее, чем при штурме на сухопутье – Russian Ministry of Defence Media Resource – April 2026. The defensible inference is that electronic warfare will remain highly effective against poorly integrated fleets, but its marginal effectiveness will decline against systems designed to operate through periods of communications denial. By 2029–2031, jamming will function less as a stand-alone kill mechanism and more as a shaping tool that reduces formation coordination, delays target recognition, forces conservative fallback behaviour and separates attack waves so that kinetic defences can engage them sequentially.

Machine vision will become the central algorithmic battlefield because both sides require automated interpretation of complex littoral scenes. Attack systems need to identify navigable corridors, avoid vessels and barriers, classify fixed infrastructure, recognise mission-relevant geometry and continue moving when communications fail. Defensive systems need to detect very small contacts, discriminate hostile behaviour, predict routes and assign effectors before the target crosses the terminal engagement line. The result will be a recursive adversarial cycle in which attackers change silhouettes, thermal profiles, paint, lighting, wake patterns and approach behaviour to manipulate classifiers, while defenders train models against synthetic variants, decoys, environmental clutter and previously unseen hulls. The US Navy’s creation of a dedicated robotics warfare specialist rating is significant because those personnel are intended to become experts in computer vision, mission autonomy, navigation autonomy, data systems, artificial intelligence and machine learning for robotic platforms—SURFOR Establishes Unmanned Surface Vessel Squadron Three – United States Navy – May 2024. In July 2026, the Naval Surface Warfare Center Panama City Division described its MEASUR project, which combines advanced sensors and artificial intelligence aboard unmanned underwater vehicles to automate assessment of physical damage to subsea objects—MEASUR: Revolutionizing Underwater Safety Through Subsea Recognition – Naval Sea Systems Command – July 2026. Although that project addresses underwater assessment rather than harbour interception, it demonstrates the institutional direction toward machine interpretation at the tactical edge. The crucial defensive weakness will be model trust. A classifier that detects ninety-nine benign objects correctly but misses one disguised attacker can fail strategically; a system that overreacts to civilian contacts may close a port, disrupt commerce or trigger an unlawful engagement. Therefore, future counter-uncrewed systems will need calibrated confidence, human-review thresholds, adversarial testing and multi-sensor corroboration rather than opaque binary outputs.

Acoustic surveillance will become increasingly important because surface and subsurface drones cannot eliminate all mechanical signatures. Propellers, waterjets, engines, pumps, cavitation, hull vibration and control-surface movements produce detectable acoustic patterns, particularly in constrained approaches where defenders can establish fixed arrays and maintain environmental baselines. Acoustic systems offer the advantage of passive operation, avoiding the electromagnetic signature generated by active radar, but performance depends heavily on water temperature, salinity, depth, seabed characteristics, shipping noise, biological activity and harbour machinery. A small electric craft may remain difficult to detect until close range, while combustion-powered systems may be identifiable by characteristic frequency components. The US Navy Marine Mammal Program officially lists threats such as explosives, mines, divers and manned or unmanned surface and subsurface vessels among objects that bottlenose dolphins and sea lions can detect or recover, illustrating that harbour security already uses non-traditional acoustic and biological sensing methods—Marine Mammal Program Supports National Security – Naval Information Warfare Center Pacific – May 2024. European work on coordinated underwater swarms further demonstrates why acoustic surveillance and communication will become inseparable parts of the offence–defence contest. The SABUVIS II programme integrated underwater vehicles and autonomous surface craft in littoral environments where communication, navigation and coordination are inherently difficult—EDA Project Develops Technology for Underwater Drones to Move in Swarms – European Defence Agency – February 2026. By 2031, advanced ports will likely combine passive hydrophones, active high-frequency sonar, magnetic sensing, optical systems and radar into a common subsurface and surface picture. Attackers will answer through quieter propulsion, speed management, drift phases, noise masking, routing near commercial traffic and expendable acoustic decoys. The defender’s advantage will be geographic familiarity; the attacker’s advantage will be the ability to choose timing, direction and signature manipulation.

Directed-energy weapons offer the most attractive theoretical answer to the defender’s cost problem because they replace expensive interceptor ammunition with electrical energy, but their practical value will vary sharply by weather, line of sight, power generation, cooling and target construction. High-energy lasers can damage sensors, ignite exposed components or weaken structural areas, while lower-power optical systems can dazzle or degrade electro-optical payloads. HELIOS, the Navy’s High Energy Laser with Integrated Optical-dazzler and Surveillance system, was developed to counter unmanned aircraft, small boats and hostile sensors while also supporting surveillance, combat identification and battle-damage assessment—NSWC Crane Collaborates to Bring High-Energy Laser Project to the Fleet – Naval Sea Systems Command – August 2021. The same Navy source emphasises speed-of-light engagement and significantly lower per-shot costs than missile-based defence, but these advantages do not create an unlimited operational magazine in practice because electrical generation, thermal management, beam quality and atmospheric attenuation constrain sustained firing. High-power microwave systems present a different mechanism: rather than heating one selected point, they can disrupt or damage electronics across a wider area, potentially making them more suitable for dense swarms. Naval Surface Warfare Center Dahlgren Division maintains separate high-energy laser and high-power microwave development divisions, reflecting the Navy’s view that the technologies have complementary roles—NSWCDD Establishes High-Powered Microwave Division – Naval Sea Systems Command – January 2022. A Chinese government defence publication in August 2025 similarly characterised high-power microwave systems as potentially useful against drone swarms and for terminal protection of ports and other critical sites, while acknowledging electromagnetic compatibility, shielding and reliability challenges—低成本、面杀伤——微波武器改写攻防规则 – Hunan Provincial National Defence Education Office – August 2025. The five-year outcome is likely to be a mixed architecture: lasers for precise sensor defeat and selected physical damage, microwaves for broader electronic disruption, and guns or interceptors for targets that survive.

Interceptor drones will emerge as the most flexible middle layer between electronic disruption and terminal gunfire. A defensive unmanned boat can patrol farther from the protected port than a fixed weapon, investigate ambiguous contacts, compel route changes, deploy entangling devices, jam locally, ram an attacker or carry its own kinetic interceptor. This shifts risk away from crewed patrol vessels and increases the depth of the defensive perimeter. The Navy’s earlier Threat Detection and Intervention System research agreement envisioned integrating autonomy into a surface vessel for unmanned threat detection and intervention, demonstrating institutional interest in defensive USV-on-USV action—USV-Based Threat Detection and Intervention System – Naval Sea Systems Command – March 2022. The European Defence Agency likewise identifies modular, redundant counter-unmanned architectures capable of targeting isolated systems and swarms from mobile land and maritime platforms—Land Domain Autonomous Systems: Counter-UAS – European Defence Agency – accessed July 2026. Although those European efforts primarily address aerial drones, the principles of distributed detection, tracking, identification and layered neutralisation transfer directly to maritime defence. The interceptor-drone layer solves two problems that fixed harbour weapons cannot. First, it creates time by engaging threats before they reach critical infrastructure. Second, it makes classification active rather than passive: a defensive craft can approach, illuminate, image, query or physically block a contact. Its limitations are equally important. Defenders must prevent capture, avoid interference with civilian traffic, manage their own communications vulnerabilities and ensure that autonomous interceptors do not become navigational hazards. By 2030, a major naval base may deploy persistent defensive flotillas in the same way that airports deploy patrols and layered air-defence systems, creating a continuous contest between attacking swarms and defending swarms rather than a one-directional attack against static infrastructure.

Five-year counter-uncrewed capability matrix

Defensive function	2026 maturity	Most likely 2031 condition	Principal attacker adaptation	Residual vulnerability
Radar and electro-optical detection	Medium–high	Automated multi-static tracking with machine-assisted classification	Lower signatures, clutter exploitation, deceptive profiles	False positives and horizon limits
Passive acoustic surveillance	Medium	Persistent fixed and mobile acoustic grids	Quieter propulsion, noise masking, drift phases	Environmental variability
Electronic warfare	High against remote control; medium against autonomy	Adaptive multi-band disruption and navigation deception	Inertial, visual and radar navigation; local mission autonomy	Spectrum congestion and fratricide
Interceptor drones	Medium	Layered surface and aerial defensive swarms	Decoys, saturation, attacks on interceptor bases	Command complexity and replenishment
Rapid-fire guns	High	Aegis-integrated automated terminal defence	Erratic manoeuvre, armour, simultaneous approach	Ammunition depletion and debris
High-energy lasers	Medium	Wider ship and shore deployment for selected targets	Reflective or sacrificial surfaces, obscurants, weather exploitation	Atmospheric attenuation and thermal limits
High-power microwaves	Low–medium	Area electronic defeat for dense swarms	Shielding, redundancy, optical or mechanical fallback	Electromagnetic compatibility and uncertain effects
Physical barriers and hardening	High	Smart barriers integrated with sensors and remote closure	Subsurface attack, breaching charges, route diversification	Fixed geometry and commercial disruption
Cyber defence	Uneven	Continuous software assurance and platform identity verification	Supply-chain compromise, malicious updates, captured hardware	Hidden dependencies and zero-day exploits

Cyber compromise will become the “shadow battlefield” because unmanned fleets contain more software, interfaces, remote updates and commercial components than conventional small craft. Attackers may seek to penetrate mission-planning systems, corrupt maps, alter geofences, steal cryptographic material, substitute sensor data, disable safety functions or implant latent vulnerabilities during manufacture and maintenance. Defenders may attempt to capture an unmanned vessel, extract software and keys, identify supplier dependencies and build exploits that can affect an entire fleet rather than one hull. The consequences of a successful intrusion are asymmetric: destroying one platform removes one asset, while compromising a common autonomy stack can undermine every vessel sharing that software. Department of Defense Directive 3000.09 requires autonomous weapon systems to incorporate cybersecurity, anti-tamper protection, verification, validation, testing and appropriate human judgment, acknowledging that software assurance is inseparable from operational safety—Autonomy in Weapon Systems, DoD Directive 3000.09 – United States Department of Defense – January 2023. The defence architecture must therefore implement secure boot, signed code, hardware-rooted identity, segmented payload control, immutable mission logs, cryptographic key rotation and the ability to isolate compromised nodes. Machine-learning systems add a separate attack surface: adversaries can poison training data, construct deceptive visual patterns, induce misclassification or exploit differences between simulation and real-world sensor behaviour. A resilient swarm must treat every shared track and every software update as potentially hostile until authenticated. By 2031, cyber defence will likely become a continuous operational function embedded at squadron level, not a periodic certification exercise. Vessels will need to prove their identity, software version and mission authority before joining the network, while commanders will require a rapid “quarantine” mechanism that prevents one anomalous platform from contaminating the wider combat cloud.

Autonomous navigation under communications denial will determine whether offensive systems can penetrate the defender’s electronic perimeter and whether defensive systems remain useful after the attacker disrupts local networks. A robust vessel cannot rely on one navigation source. It must combine inertial measurement, compass data, radar, electro-optical imagery, depth information, stored maps, speed estimates and possibly celestial or coastline references. The system must understand uncertainty rather than simply output a location, because errors accumulate when satellite navigation disappears. A vessel that knows its position only within a broad confidence ellipse may remain capable of reaching a harbour entrance but not of safely identifying a specific berth. The operational response will be hierarchical autonomy: strategic objectives and engagement constraints remain human-defined, while local software executes route replanning, collision avoidance and formation maintenance. NIWC Atlantic’s Vesuvius Spiral 2 experiment demonstrated remote control of unmanned vessels in Portugal from a mission centre in Naples and validated the management of multiple heterogeneous systems by a single operator, highlighting both the promise and the dependence of distributed operations on reliable data architecture—Naval Engineers Control Unmanned Vessels Across Countries – Naval Information Warfare Center Atlantic – June 2025. The next phase must prove not merely that remote control works over distance, but that the mission continues safely when the long-distance link disappears. European mixed-swarm research beneath the surface, where communications are inherently constrained, reinforces the same development path toward local task allocation and cooperative navigation—EDA Project Develops Technology for Underwater Drones to Move in Swarms – European Defence Agency – February 2026. The defender will therefore seek not only to deny communications but to increase the attacker’s navigational uncertainty until autonomous safety logic causes delay, route abandonment or misallocation of forces.

Industrial attrition will ultimately decide the competition because offence and defence both require replenishment at a rate compatible with sustained operations. A fleet of inexpensive attack drones can impose disproportionate costs if the defender responds with high-end missiles, but the advantage disappears when the defender employs guns, reusable interceptors, lasers, microwave systems or barriers whose marginal engagement cost is lower. Conversely, defensive systems require around-the-clock crews, maintenance, power, spare sensors and ammunition, while attackers can concentrate production and launch only when conditions are favourable. The United States Navy now reports owning hundreds of small unmanned surface vessels, compared with approximately four a year earlier, indicating that scale is moving from aspiration toward operational inventory—SNA 2026: Status of the Surface Force – Commander, Naval Surface Force, US Pacific Fleet – January 2026. Yet the Government Accountability Office warned in June 2026 that the Navy still faces leadership, organisational and fielding challenges in robotic and autonomous systems—Robotic Autonomous Systems: Navy Needs to Address Leadership and Organizational Challenges – US Government Accountability Office – June 2026. GAO has also found that earlier Navy planning did not fully incorporate operations, sustainment and digital-infrastructure costs into estimates for uncrewed maritime systems—Navy Shipbuilding: A Generational Imperative for Systemic Change – US Government Accountability Office – March 2025. Industrial power therefore cannot be measured solely by hull output. The relevant variables are mission-capable availability, repair time, component commonality, software release speed, engine supply, sensor replacement, launch infrastructure and the ability to regenerate trained operators. A force that produces three hundred vessels but cannot sustain their communications, payloads or maintenance cycle possesses inventory, not enduring combat power.

Competing hypotheses and Bayesian update

Hypothesis	Proposition	Evidence supporting it	Evidence challenging it	Updated probability
H₁ — Offensive Saturation Dominance	Massed autonomous boats overwhelm any affordable defence	Low-cost hulls, multiple axes, defender identification burden	Layered guns, EW, barriers and interceptors	22%
H₂ — Electronic-Warfare Defence Dominance	Communications and navigation disruption neutralise most attacks	Remote-control dependency remains common	Machine vision and autonomous navigation reduce dependence	16%
H₃ — Layered Harbour Defence Equilibrium	Integrated sensing and mixed effectors contain attacks without eliminating them	Existing sensor, gun, barrier and EW technologies	High cost of continuous coverage and false positives	27%
H₄ — Directed-Energy Cost Reversal	Lasers and microwaves restore a decisive defender cost advantage	Low marginal shot cost and rapid engagement	Weather, power, cooling and shielding constraints	14%
H₅ — Software and Cyber Primacy	The side controlling code, data integrity and autonomy assurance gains the decisive edge	Common software affects entire fleets; machine vision centrality	Cyber effects are difficult to guarantee under combat conditions	21%

The updated assessment assigns the highest probability to H₃, a layered but unstable equilibrium in which no single defensive technology eliminates the unmanned maritime threat. The prior distribution began at 20% for each hypothesis. Verified fielding of Aegis-integrated 30 mm defence, expanding autonomous inventories and European counter-unmanned research raised the probability that defenders will construct effective layered systems. Evidence of machine vision and increasing autonomy reduced the probability of pure electronic-warfare dominance. Directed energy retains substantial long-term potential, but atmospheric conditions, power generation and technical maturity constrain its probability of becoming the sole cost-reversing solution by 2031. Software and cyber primacy remain highly plausible because both offensive and defensive systems rely on common autonomy stacks, sensor fusion and networked command; however, cyber effects are difficult to reproduce reliably against hardened systems and may require access that cannot be assumed. A Monte Carlo-style scenario model using these hypotheses produces three broad outcomes: a 44% probability of contested equilibrium, in which attacks remain possible but require increasingly sophisticated coordination; a 32% probability of temporary offensive advantage, driven by mass, surprise or defender saturation; and a 24% probability of defensive cost reversal in selected high-value ports equipped with mature directed energy, interceptors and sensor fusion. These values are structured analytical estimates rather than official statistics. Their principal implication is that geography will produce uneven results. Major naval bases may become extremely difficult to penetrate, while secondary ports, logistics anchorages, offshore infrastructure and commercial chokepoints remain exposed because they cannot justify equivalent defensive density.

The five-year competition will therefore be governed by adaptation speed rather than static technical superiority. Between 2026 and 2027, defenders will accelerate integration of radar, electro-optical sensors, machine-assisted classification, rapid-fire guns and electronic warfare into existing combat systems. During 2027–2028, interceptor drones and remote patrol craft will extend defensive depth, while attackers introduce more autonomous navigation and multi-axis deception. During 2028–2029, the contest will shift toward adversarial machine vision, track manipulation, acoustic signature management and cyber attacks against common software. During 2029–2030, high-energy lasers and microwave systems are likely to assume a larger role at selected naval bases and major ships, although not as universal replacements for kinetic weapons. By 2030–2031, mature defensive networks will combine physical barriers, passive sensors, active radar, acoustic arrays, electronic disruption, autonomous interceptors, directed energy and guns under a unified command layer. Attackers will respond with heterogeneous swarms containing reconnaissance craft, communications relays, decoys, electronic-warfare nodes, surface attackers, aerial drones and possibly subsurface vehicles. The resulting battlefield will resemble a distributed siege conducted through software and machines. Success will depend on who can observe, classify, decide, regenerate and adapt faster—not merely who possesses the fastest boat or most powerful laser. The central strategic risk is that automation compresses decision time while increasing uncertainty. A defender facing dozens of ambiguous contacts may delegate progressively more classification and engagement functions to machines, while an attacker may pre-authorise autonomous continuation after communications loss. That interaction creates the possibility that tactical systems execute correctly according to their local rules while producing an escalation neither political leadership intended nor can immediately stop.

Pillar III — Strategic, Legal and Geoeconomic Consequences

The strategic consequence of distributed maritime autonomy is not simply that states acquire another class of precision weapon. It is that they gain a scalable instrument for exerting pressure below, around and potentially across the threshold of conventional war. A crewed naval attack normally carries conspicuous preparatory signatures: ships deploy, personnel are placed at risk, rules of engagement are elevated, logistics movements become visible and political leaders assume responsibility for possible casualties or capture. Uncrewed maritime systems can reduce several of those constraints. Small autonomous or remotely operated vessels can be transported commercially, launched from dispersed coastal locations, pre-positioned aboard auxiliary ships, activated in stages and exposed to loss without creating an immediate personnel-recovery crisis. This lowers the operational cost of testing an opponent’s surveillance, disrupting port activity, threatening offshore infrastructure or imposing temporary exclusion zones. It may also lower the perceived political cost of initiating force because the attacker does not immediately risk sailors aboard the platform. That reduction in friendly exposure does not make coercion casualty-free or legally neutral; it changes the distribution of risk by moving danger from the attacking crew toward defenders, civilian mariners, port workers, coastal populations and commercial infrastructure. The resulting escalation geometry is unstable because an unmanned incursion may be interpreted as reconnaissance, sabotage, deception, preparation for a larger strike or an already authorised lethal attack. Political leaders may have only minutes to decide whether an approaching vessel constitutes an armed attack, an intelligence probe or a navigational accident. UNCTAD has already warned that disturbances at strategic chokepoints transmit shocks across supply chains, commodity markets and development prospects, and its March 2026 assessment of the Strait of Hormuz explicitly linked maritime disruption to the need to protect ports, seafarers, civilian infrastructure and secure trade corridors—Strait of Hormuz Disruptions: Implications for Global Trade and Development – United Nations Conference on Trade and Development – March 2026. The strategic novelty is therefore the combination of low personnel exposure, ambiguous attribution, relatively low platform replacement cost and potentially high economic leverage.

Escalation control becomes more difficult as command authority is distributed across software, remote operators, mission commanders, intelligence systems and autonomous fallback rules. In a crewed vessel, the chain of command remains physically concentrated, and the commanding officer can observe local circumstances, interpret intent and abort action. In a distributed swarm, different components may possess different mission data, communications quality and degrees of autonomy. One node may retain contact with command, another may continue under a pre-authorised route, and a third may misclassify defensive manoeuvres as hostile interference. If the swarm was programmed to preserve mission effectiveness after communications denial, a political decision to halt the operation may not propagate to every platform. The key escalation variable is therefore not nominal autonomy but reversibility: whether political and military authorities can reliably suspend, redirect or terminate distributed action after launch. This must be designed into communications architecture, mission logic and legal approval rather than assumed. The United Nations General Assembly affirmed in December 2025 that international law, including the UN Charter, international humanitarian law and international human rights law, applies to autonomous weapons systems, while recognising serious legal, security, technological and ethical concerns—Lethal Autonomous Weapons Systems, Resolution A/RES/80/57 – United Nations General Assembly – December 2025. The UN Secretary-General has separately called for a legally binding instrument prohibiting systems operating without human control, underscoring that political concern now centres on retaining meaningful authority over life-and-death decisions—Secretary-General’s Remarks to the Security Council on Artificial Intelligence and International Peace and Security – United Nations – September 2025. In maritime operations, effective escalation control will require abort channels, mission-expiry timers, geographically bounded authorities, degraded-mode constraints, independent authentication and an auditable hierarchy showing which human approved each transition from surveillance to interference, from interference to disabling action and from disabling action to lethal force.

Human accountability will remain legally indispensable even when machines perform navigation, classification, target tracking and terminal manoeuvre. A vessel cannot bear criminal responsibility, exercise command responsibility, compensate victims or explain why a particular target was selected. Responsibility therefore remains attached to humans and institutions, but autonomy can diffuse the evidentiary trail across designers, manufacturers, data suppliers, software integrators, commanders, intelligence analysts and operators. This creates a risk of the “many hands” problem: each participant controls only one component and may argue that the harmful outcome emerged from the system as a whole. The legal answer cannot be to treat machine behaviour as unforeseeable whenever an autonomous platform produces an unintended result. Commanders remain responsible for selecting means and methods of warfare, approving operational environments, assessing foreseeable failure modes and ensuring compliance with distinction, proportionality and precaution. Developers and procurement authorities may incur other forms of responsibility where negligent testing, undisclosed defects, inadequate cybersecurity or unsafe configuration contribute to harm. The International Committee of the Red Cross has consistently framed autonomous weapons as systems that raise questions about human control, predictability and accountability, while international humanitarian law continues to regulate those who design, authorise and employ them—Autonomous Weapon Systems – International Committee of the Red Cross – official thematic resource. The United Nations Security Council’s September 2025 debate record also included the position that meaningful human control is necessary for legal compliance and achievable accountability—Security Council Meeting Record S/PV.10005 – United Nations – September 2025. For maritime autonomy, accountability must consequently become a systems-engineering requirement. Mission logs should preserve command orders, time stamps, model versions, sensor inputs, classification confidence, communication failures, geofence changes and operator interventions. Without tamper-resistant records, post-strike investigations may be unable to distinguish deliberate conduct from technical malfunction, cyber manipulation or foreseeable design failure.

Attribution constitutes a separate strategic problem because identifying the manufacturer or apparent origin of an unmanned vessel does not prove who launched, controlled or authorised it. Commercial engines, satellite terminals, cameras, navigation units and composite hull materials circulate across borders. Software may be modified, identifiers removed and components deliberately selected to implicate another actor. A state may transfer platforms to a proxy while retaining intelligence support, mission planning or remote override. Conversely, an adversary may capture a system and redeploy it as a false-flag instrument. Attribution therefore requires a layered evidentiary model combining physical forensics, communications intelligence, satellite observation, financial records, logistics movements, software artefacts, launch-site surveillance and analysis of who benefited from the operation. No single indicator should be treated as dispositive. The proliferation of common autonomy stacks and commercially sourced components will make attribution slower precisely when political leaders feel pressure to respond rapidly. This creates a dangerous asymmetry: the tactical event may unfold in minutes, while high-confidence attribution may require days or weeks. Retaliation before confidence matures can strike the wrong actor; waiting can invite repeated attacks and weaken deterrence. The appropriate architecture is a graded attribution framework in which states distinguish technical attribution, operational attribution, organisational attribution and state responsibility. Technical attribution asks how the system functioned; operational attribution asks where it came from and how it was controlled; organisational attribution asks which military, intelligence service, proxy or company executed the mission; legal attribution asks whether conduct is attributable to a state under international law. Chinese official military commentary has openly discussed unmanned swarm structures involving different classes of unmanned boats conducting autonomous search, identification, escort and interception, demonstrating that replication of distributed maritime concepts is not confined to Western forces—作战无人蜂群：振翅欲飞知向谁边 – Ministry of National Defense of the People’s Republic of China – April 2020. As more states and proxies acquire similar systems, attribution will increasingly depend on intelligence integration rather than visual identification.

Insurance markets will become an early commercial sensor of autonomous-maritime risk because underwriters must translate uncertainty into premiums, exclusions, deductibles and coverage conditions before governments finish negotiating comprehensive rules. Autonomy may reduce some conventional losses by decreasing fatigue, navigational error, piracy exposure and crew-related claims, but it creates new concentrations of risk in software, remote-control centres, satellite communications, cyber dependencies and shared vendors. A defect in one navigation algorithm could affect an entire fleet simultaneously, transforming what was once an isolated casualty into a correlated portfolio event. Marine insurers will need to determine whether a collision resulted from vessel programming, negligent remote supervision, defective charts, communications failure, cyberattack, adversarial spoofing or force majeure. War-risk insurers will face additional ambiguity when an ostensibly commercial autonomous vessel is requisitioned, remotely redirected or misidentified during hostilities. UNCTAD previously identified safety, cybersecurity, liability and insurance as principal concerns arising from autonomous ships and drones—Review of Maritime Transport 2018, Chapter V – United Nations Conference on Trade and Development – 2018. It also noted that autonomy could reduce certain human-error and piracy-related losses while creating unresolved questions over responsibility and compensation—Review of Maritime Transport 2019, Chapter IV – United Nations Conference on Trade and Development – 2019. By 2031, insurance coverage is likely to depend on demonstrable compliance with technical-assurance conditions: certified software, redundant communications, cyber-risk controls, incident logging, remote-operator qualifications, geofencing, independent testing and clear contractual allocation of responsibility. Ports may also require proof that autonomous vessels carry valid liability coverage and can identify a legally responsible operator before granting entry. The market will therefore create de facto standards even where public law remains incomplete. Insurers may refuse coverage or impose punitive premiums on opaque autonomy architectures, unsupported software or vessels unable to prove who held operational control at the time of an incident.

Port security will require a fundamental revision because the existing regime was designed primarily around terrorism, unauthorised access, cargo security, ship–port interfaces and human-controlled vessels. The International Ship and Port Facility Security Code, mandatory under SOLAS Chapter XI-2 since July 2004, establishes a risk-management framework for assessing threats and determining appropriate protective measures—SOLAS XI-2 and the ISPS Code – International Maritime Organization – official security framework. The IMO–ILO Code of Practice on Security in Ports extends that perspective beyond individual port facilities to the wider port environment—Code of Practice on Security in Ports – International Maritime Organization and International Labour Organization. Autonomous threats expose gaps in both frameworks because a hostile system may enter without a crew, forged documentation or conventional boarding indicators. Port-facility security assessments will need to include underwater approaches, autonomous loitering zones, unmanned launch points, navigation-data compromise, remotely activated payloads and attacks on digital port systems that facilitate physical penetration. Security levels may need dynamic machine-readable implementation so that autonomous commercial vessels automatically receive altered routing, speed restrictions or exclusion instructions during elevated alerts. Ports must also differentiate authorised industrial robots, survey craft, pilot-assistance systems and logistics drones from hostile platforms. This requires identity credentials resistant to spoofing, authenticated machine-to-port communications and procedures for vessels that cannot establish trusted identity. By 2031, high-value ports will likely establish autonomous-vessel control zones analogous to controlled airspace, with compulsory reporting, approved corridors, continuous telemetry and remote inspection authority. Less-resourced ports may lack the capital and personnel to implement such systems, creating a two-tier maritime economy in which secure hubs attract traffic and insurance capacity while vulnerable ports face higher costs, operational restrictions or avoidance.

Naval force structure will change not through immediate replacement of crewed warships but through redistribution of tasks, risk and magazine depth. Large crewed ships will remain necessary for command, air defence, aviation, endurance, heavy weapons, repair and political presence. Uncrewed systems will increasingly occupy the outer layers of the force: reconnaissance, decoy operations, electronic surveillance, mine warfare, communications relay, logistics, submarine tracking and high-risk penetration. This can allow scarce crewed ships to remain farther from shore or concentrate on missions requiring human judgement and substantial payload. The danger is that navies may count unmanned hulls as substitutes for conventional capacity without funding the control networks, maintenance infrastructure and support vessels that make them operational. The US Government Accountability Office reported in June 2026 that the Navy still needed portfolio management, clearer stakeholder responsibilities and more effective development processes to field robotic autonomous systems rapidly—Robotic Autonomous Systems: Navy Needs to Address Leadership and Organizational Challenges to Meet Urgent Needs – US Government Accountability Office – June 2026. GAO had already warned that a planned $4.3 billion investment in 21 robotic autonomous maritime systems did not fully account for operations, sustainment and digital infrastructure, creating affordability risk—Navy Shipbuilding: A Generational Imperative for Systemic Change – US Government Accountability Office – March 2025. This indicates that force structure must be measured in operational ecosystems rather than hull totals. A credible autonomous squadron includes command nodes, technicians, software teams, communications capacity, transport, launch systems, spares, test equipment and trained legal and intelligence personnel. The five-year outlook therefore favours hybrid fleets: crewed platforms providing command, protection and heavy effects; uncrewed systems expanding reach and accepting attrition; and shore-based networks supplying data, maintenance and mission updates.

Industrial mobilisation will move naval competition away from traditional shipyard capacity alone and toward a broader manufacturing ecosystem involving commercial boat builders, electronics producers, software companies, battery suppliers, composite manufacturers and telecommunications providers. This creates both resilience and vulnerability. Small autonomous craft can be produced in facilities that cannot construct destroyers or submarines, potentially allowing rapid scaling and geographic dispersion. Commercial components can shorten development time and reduce costs. However, dependence on globally sourced processors, cameras, motors, satellite terminals and navigation units creates exposure to export restrictions, sabotage, counterfeit parts and supplier concentration. The conventional shipbuilding base already faces structural capacity and workforce constraints. GAO found in February 2025 that the US shipbuilding and repair industrial base faced challenges supporting Navy goals and that the Navy lacked a fully strategic approach to managing the sector—Shipbuilding and Repair: Navy Needs a Strategic Approach for Its Industrial Base – US Government Accountability Office – February 2025. Autonomous systems may relieve some pressure by shifting work to smaller manufacturers, but they also introduce new bottlenecks in secure electronics, software certification and integration testing. Industrial mobilisation must therefore be assessed across five dimensions: production throughput, component security, software-update capacity, repairability and combat replacement. The decisive metric is not how many vessels a factory can launch in peacetime but how quickly a force can replace losses while preserving trusted software, payload availability and communications access under sanctions or attack. By 2031, leading states are likely to maintain surge contracts, pre-qualified component alternatives, modular mission packages and digital designs that can be produced across multiple facilities. States unable to secure critical electronics or develop trusted autonomy software may still manufacture hulls but remain dependent on external suppliers for the decisive layer of capability.

Adversary replication and proliferation will occur rapidly because many enabling technologies are dual-use and widely available. Commercial autonomy for offshore inspection, hydrographic survey, environmental monitoring and port logistics already relies on sensors, navigation software, satellite communications and small-vessel control systems that can be adapted for military purposes. The barrier to entry is lower than for submarines, combat aircraft or precision missiles, although reliable long-range navigation, swarm coordination and target discrimination remain technically demanding. Chinese official defence sources have described unmanned boats conducting reconnaissance, patrol, interception and collaborative swarm missions, illustrating that major competitors already possess conceptual frameworks for replication—未来智能无人艇的应用前景 – Ministry of National Defense of the People’s Republic of China – June 2021. The more dangerous proliferation pathway may involve non-state armed groups that do not require military-grade reliability. A proxy may accept high failure rates because the systems are inexpensive, deniable and useful for forcing port closures or raising insurance premiums even when they fail to hit a target. Commercially available components also permit distributed assembly, complicating sanctions and interdiction. Proliferation control will therefore need to focus not only on complete vessels but on high-risk combinations of autonomy software, long-range communications, payload interfaces, advanced navigation and swarm-management tools. Conventional export-control lists may struggle because each component has legitimate civilian uses. States may instead regulate technical assistance, software access, mission-specific integration and bulk procurement patterns. The likelihood of successful control remains limited: once operational concepts are demonstrated publicly, replication can proceed through experimentation without direct technology transfer. The most realistic policy objective is therefore not preventing all diffusion, but increasing the cost of acquiring resilient, coordinated and accurately targeted systems while strengthening defensive adaptation and forensic attribution.

Strategic consequence matrix, 2026–2031

Consequence vector	Near-term effect	Five-year structural outcome	Principal risk indicator	Net strategic direction
Escalation control	Reduced political cost of risking platforms	Faster crises with weaker recallability	Loss of communications after mission launch	Negative
Human accountability	Responsibility distributed across technical and command chains	Mandatory auditability and clearer human authorisation requirements	Missing or non-verifiable mission logs	Contested
Attribution	Component origin mistaken for operator identity	Multi-domain forensic attribution becomes essential	Use of common commercial hardware	Negative
Insurance	Premium uncertainty and new cyber exclusions	Assurance-based underwriting and autonomous-vessel conditions	Common-mode software failures	Mixed
Port security	Existing perimeter models become inadequate	Autonomous control zones and persistent multi-sensor surveillance	Unidentified low-signature contacts	Cost-increasing
Naval force structure	Uncrewed systems augment high-risk missions	Hybrid fleets with distributed outer layers	Unfunded support and digital infrastructure	Transformative
Industrial mobilisation	New entrants expand hull production	Competition shifts toward electronics, software and regeneration	Single-source critical components	Strategically decisive
Proliferation	Dual-use systems lower barriers	State and proxy replication becomes widespread	Commercial procurement anomalies	Strongly negative
Maritime coercion	Cheap platforms threaten costly infrastructure	More frequent below-threshold pressure on ports and chokepoints	Repeated ambiguous incursions	Destabilising

The changing cost of maritime coercion will be the most consequential geoeconomic effect. Traditionally, coercing maritime trade required submarines, mines, anti-ship missiles, aircraft, major surface forces or physical seizure of vessels. Autonomous systems add a cheaper method for creating uncertainty without necessarily destroying large amounts of infrastructure. A small number of credible incidents can force shipping companies to reroute, slow approaches, suspend port calls, hire additional security or accept higher insurance premiums. The economic effect may substantially exceed the replacement cost of the attacking systems. UNCTAD’s Review of Maritime Transport 2025 describes elevated and volatile freight rates, chronic port disruption, route diversion and growing exposure to geopolitical tension—Review of Maritime Transport 2025 – United Nations Conference on Trade and Development – September 2025. Its freight-cost analysis notes that tensions around the Strait of Hormuz add risk to regional hubs and maritime chokepoints, even where the direct share of container trade through the strait is smaller than its importance to energy flows—Freight Rates and Maritime Transport Costs, Review of Maritime Transport 2025 – United Nations Conference on Trade and Development – September 2025. Unmanned maritime coercion can therefore function as an economic weapon of expectation: the attacker does not need to sink every ship, only convince market participants that safe passage is uncertain. This creates a leverage ratio in which low-cost tactical actions generate higher insurance, inventory, fuel and financing costs across entire supply chains. The defender must spend continuously to guarantee security, whereas the attacker may spend intermittently to disrupt confidence. By 2031, the cost of protecting a major port may include persistent autonomous patrols, integrated sensor networks, cyber operations centres, physical barriers, directed-energy systems and insurance-backed compliance regimes. Maritime coercion will consequently become more accessible but not necessarily more decisive; its effectiveness will depend on whether defenders can restore commercial confidence faster than attackers can regenerate uncertainty.

Analysis of competing hypotheses and Bayesian update

Hypothesis	Core proposition	Supporting evidence	Contradictory evidence	Updated probability
H₁ — Controlled Deterrence	Autonomous fleets strengthen deterrence without materially increasing escalation	Reduced friendly exposure; improved surveillance and defensive persistence	Ambiguous intent and communications loss compress decision time	15%
H₂ — Persistent Grey-Zone Coercion	States and proxies use unmanned systems for deniable pressure below conventional-war thresholds	Low platform cost, dual-use components and attribution difficulty	Improved forensics and defensive identification	29%
H₃ — Regulatory Stabilisation	Law, insurance and technical standards constrain unsafe deployment	IMO MASS Code, UN negotiations and assurance requirements	Military innovation outpaces civilian regulation	16%
H₄ — Hybrid Naval Transformation	Uncrewed systems permanently reshape force structure while remaining subordinate to human command	Navy portfolio investment and hybrid operational concepts	Sustainment and digital-infrastructure deficits	24%
H₅ — Proliferation-Driven Instability	Replication by regional powers and proxies drives recurrent port and chokepoint crises	Dual-use supply chains and official multinational swarm development	Reliable long-range autonomy remains difficult	16%

The posterior distribution places the greatest weight on H₂, persistent grey-zone coercion, and H₄, hybrid naval transformation. The model begins with equal 20% priors and updates against nine evidence categories: declining platform-risk cost, attribution difficulty, commercial component availability, existing international law, developing maritime regulation, industrial constraints, port vulnerability, insurance response and the continued need for human command. The probability of controlled deterrence falls because autonomous systems make probing and limited force easier while creating uncertainty over intent and recallability. Regulatory stabilisation remains credible but slower than military adoption; the non-mandatory MASS Code entered effect on 1 July 2026, while the IMO roadmap anticipates a mandatory code by 1 July 2030 and entry into force in 2032, leaving most of the five-year analytical period governed by an evolving rather than settled framework—IMO Adopts First Global Code for Autonomous Ships – International Maritime Organization – May 2026. The proliferation hypothesis remains significant but does not dominate because sophisticated, resilient maritime swarms still require systems integration, testing and command architecture beyond basic commercial conversion. A Monte Carlo scenario model using 50,000 illustrative iterations and uncertainty bands for regulatory maturity, proxy access, insurance sensitivity, port-defence investment and attribution confidence yields a 46% probability of recurrent but bounded coercive incidents, a 27% probability of accelerated hybrid-fleet transformation without major systemic crisis, a 17% probability of proliferation-driven regional escalation and a 10% probability of strong regulatory and defensive adaptation substantially containing the risk. These are structured analytical estimates, not observed frequencies. The strongest warning indicator is the convergence of proxy access, reliable autonomous navigation and ambiguous commercial signatures.

The legal and strategic architecture required by 2031 is therefore a layered governance system rather than a single treaty or technical standard. At the operational level, forces need explicit human authorisation thresholds, positive-control mechanisms, mission-expiry rules and autonomous fallback behaviours designed to minimise escalation. At the technical level, systems need tamper-resistant logging, authenticated software, secure identities and the ability to reconstruct why a platform acted. At the national level, procurement authorities need legal reviews, independent testing, cyber certification and clear allocation of responsibility among commanders, manufacturers and software providers. At the international level, states need norms for identification, distress response, navigation, interaction with civilian vessels and investigation of incidents involving autonomous platforms. The IMO MASS Code, effective from July 2026 for applicable cargo ships on a non-mandatory basis, establishes a safety-oriented framework for operational modes, connectivity, remote operations, security, navigation and the human element—Autonomous Shipping FAQ – International Maritime Organization – July 2026. It does not regulate naval combat systems, but it will shape expectations concerning safe operation and demonstrable control. Military autonomy will remain governed principally by the law of armed conflict, national weapons reviews and rules of engagement. The danger is regulatory fragmentation: civilian autonomous vessels may carry visible identities and certified safety systems, while military and proxy systems deliberately exploit anonymity. States should therefore avoid equating compliance mechanisms with universal protection. The systems most likely to threaten ports may be precisely those outside formal regulatory regimes. Effective governance must combine law with surveillance, intelligence, industrial control and defensive resilience.

The ultimate five-year judgement is that autonomous maritime systems will reduce the unit cost of presence, surveillance and attack while increasing the aggregate cost of maritime security, legal assurance and commercial risk management. They will not abolish large warships, human command or traditional deterrence. They will surround those institutions with a faster, more ambiguous and more replaceable layer of machine activity. Escalation will become harder to manage because platforms may continue after links fail, incidents may be difficult to attribute and tactical losses will impose less political restraint on the attacker. Accountability will become more technically demanding because lawful review depends on preserving complex digital records. Insurance and port-access conditions will become instruments of governance, imposing market discipline where public regulation remains incomplete. Naval forces will become hybrid, but their effectiveness will depend on networks, logistics and software rather than impressive hull counts. Industrial mobilisation will favour states capable of combining commercial manufacturing with trusted electronics and rapid software iteration. Proliferation will expand because dual-use components are accessible, although advanced coordinated autonomy will remain unevenly distributed. The cost of maritime coercion will fall for attackers able to create uncertainty, while the cost of credible protection will rise for ports, shipping companies and coastal states. This imbalance does not guarantee offensive dominance: layered defence, strong attribution and rapid restoration of commercial confidence can reduce coercive leverage. It does mean that maritime power must increasingly be measured not only by ships and missiles but by the capacity to govern, insure, identify, secure, regenerate and legally control distributed autonomous forces.

---

Copyright of debugliesintel.com
Even partial reproduction of the contents is not permitted without prior authorization – Reproduction reserved