In spring 2026, Britain, Norway, and the United States tracked and disrupted a covert Arctic exercise off the coast of Norway involving Russian deep-sea submersibles simulating the deployment of a weapon designed to disable the two 1,400-kilometer submarine cables linking the islands of Svalbard to mainland Norway. Svalbard hosts a satellite ground station serving Europe, the United States, and Japan. This encounter revealed that NATO allies can detect a covert subsea operation and, if caught in time, disrupt it.
Had it been a real attack, the first question would have been where and how quickly internet and sensitive government traffic could be rerouted. But this requires grappling with who will repair the cables, how quickly, and what would happen if additional cables in other theaters were cut before the first repair was complete. To effectively disrupt global communications, an adversary needs only to damage cables in multiple locations faster than the repair system can restore them.
New cable routes being developed to avoid geopolitical choke points face additional restoration challenges. Nordic research networks are planning a 10,000-kilometer Polar Connect cable from Sweden across the Arctic Ocean to Japan and North America, a route intended to provide an alternative path to the Red Sea. Its planned route runs close to Svalbard. The route will pass through the same waters where Russia recently rehearsed disrupting cables.
In 2017, Adm. (ret.) James Stavridis warned in the foreword to a Policy Exchange report by future U.K. Prime Minister Rishi Sunak, “It is not satellites in the sky, but pipes on the ocean floor that form the backbone of the world’s economy. … We have allowed this vital infrastructure to grow increasingly vulnerable and this should worry us all.” Since 2020, the Federal Communications Commission has implemented strong policies around submarine cable ownership and supplier screening. But the United States has no wartime repair and restoration strategy for a campaign of simultaneous cable failures across multiple theaters.
Wartime submarine cable repair should be treated as a national security mission and funded as such. The designation of submarine cables as critical infrastructure, the establishment of a federal coordinating authority with emergency powers to prioritize repairs, and an increased budget to expand the Cable Security Fleet would give the United States a way to decide which cables are repaired first and which ships are needed to carry out those decisions.
Vulnerability in the Repair Queue
Cable faults are routine. Fishing gear, anchors, and other accidents damage submarine cables roughly 150 to 200 times annually, a number easily handled by the approximately 22 ships in the world equipped to repair subsea cables. Under normal conditions, this system works well; while one cable is under repair, operators reroute traffic over spare capacity on other paths. But several faults at once can exhaust that spare capacity and clog the queue for repair ships. The Hunga Tonga–Hunga Ha’apai eruption of 2022 caused an underwater landslide that isolated the island nation of Tonga for five weeks, with full restoration taking almost 18 months. And in February 2023, five of Vietnam’s international cables were damaged within a short time frame, with full restoration happening nine months later.
Repair capability also depends on several factors: access: ship location, its flag, who legally controls it, a trained crew, whether it can be insured in a conflict zone, and the availability of spare cable and parts. The Pacific illustrates the risk. Repairs there are coordinated through the Yokohama Zone maintenance agreement, which relies on three operators, including S.B. Submarine Systems, a Shanghai-based cable installation and maintenance company that is majority owned by China Communications Services, whose controlling shareholder is state-owned China Telecommunications Corporation. It has repaired cables owned by Google and Meta. U.S. officials have raised security concerns about American dependence on Chinese repair vessels. In a conflict involving China or Taiwan, the United States could not safely assume continued access to Chinese-controlled repair capacity. Beijing would not need to sabotage anything. It could simply decline to sail and repair.
The United States doesn’t have a government-operated repair ship, since the Navy cable ship Zeus was officially deactivated in 2025 after four decades of service. Instead, the Cable Security Fleet is authorized at $10 million annually for payments to the operator of two privately owned ships to make them available to the Department of Defense when requested. This provides a minimum assured capability for repairing an isolated failure but does little for simultaneous cuts across multiple theaters. A contract with two ships is not the same as having two ships immediately available to sail. Since they operate commercially, these ships can be contracted to work far from their assigned ocean. For example, in 2022, the cable ship Dependable spent a month in the Indian Ocean laying a cable to the Naval Support Facility in Diego Garcia, far from its coverage area of the Pacific.
Less discussed is the criticality of the subsea cable supply chain. A repair ship is only as useful as the spares it carries: jointing kits, repeaters, and cable that matches the system being repaired. These components are not interchangeable and not always readily available. Most of Tonga’s 18-month wait went to confirming specifications, then manufacturing and shipping the right replacement cable from France. U.S. policy is also narrowing the supplier pool. In 2025, the Federal Communications Commission barred equipment on its Covered List of national security risks from U.S.-connected cable systems, and the Commission’s Third Report and Order banned devices containing Covered List components. A wartime restoration strategy therefore requires stockpiling trusted spares before they are needed.
Europe has begun addressing this problem. In 2026, the European Union allocated €60 million (or about $70 million — €20 million for the Baltics, €40 million for the Mediterranean and Atlantic) for modular cable repair equipment to be pre-positioned at ports and shipyards. These can be rapidly installed on vessels pre-configured to receive the module. The United States has no equivalent.
Repair Under Fire
In armed conflict, a vessel or cable can become a lawful military objective if the targeting criteria are met. The Second Geneva Convention protects hospital ships but says nothing about cable ships, whose legal and operational risk can change depending on how they are used during hostilities. The 2025 Newport Manual on the Law of Naval Warfare treats military communications cables as military objectives. However, in practice, it is difficult to distinguish military communications cables from commercial ones because military and government communications often travel over commercial cable systems.
In a 1923 arbitration, Cuba Submarine Telegraph Company (Great Britain) v. United States, a tribunal rejected a British company’s claim for cables the U.S. Navy cut in Cuban waters during the Spanish-American War, finding the company functioned as a Spanish public service. The same logic could apply today to a commercial cable that carries military traffic. Cable repair ships are essential to restoring both military and civilian communications but could become military targets because they have no immunity from attack. As Douglas Burnett, former legal adviser to the International Cable Protection Committee, has argued, cables and repair ships remain legitimate targets in a conflict.
Insurance compounds the problem, because war coverage can be canceled on as little as 72 hours’ notice. After U.S. hostilities with Iran began on Feb. 28, 2026, major insurers issued cancellation notices for war-risk cover for ships in the Persian Gulf within days. By early March, seafarer unions and employers designated the Strait of Hormuz, the Gulf of Oman, and the Persian Gulf a Warlike Operations Area, giving crews the right to refuse to sail there. Although some insurers brought back limited, expensive war-risk coverage, the Congressional Research Service noted that ship owners were still reluctant to put crews in harm’s way regardless of whether insurance was available.
Repair and restoration in contested waters cannot be treated as a purely commercial service. For example, in the 2024 Red Sea incident in which Houthi rebels struck the Rubymar and dragged its anchor across three cables, full repairs took about six months, delayed by security concerns and a political dispute in Yemen over who could grant repair permits. Repair situations such as this cross commercial, civilian, and military authorities, which complicates operations. complicating operations.
This problem is exacerbated if a cut is deliberate, because the repair ship could become a target. Since cables carry military and civilian traffic, it is unclear whether a repair ship serves a military or a civilian purpose. A naval escort would settle that question, but not in the ship’s favor. Under the law of naval warfare, a merchant ship sailing under convoy of a belligerent’s warships may be treated by the other side as a military objective. For a commercial ship repairing commercial cables, that cost is real: without an escort, it might keep its civilian status. For the ships the United States would rely on most, the cost is smaller. A ship under a Department of Defense contract restoring cables that carry military traffic is likely already a military objective, escorted or not, because it is contributing to military action. Below the level of armed conflict, an escort deters interference without the legal risk. If crews and insurers will not sail unprotected, the alternative to an escorted repair might be no repair at all. Protection should therefore depend on whose ship it is and what it is repairing: close escort for ships under government contract, and area security at a distance for purely commercial ones.
The Peacetime Record Is Misleading
The strongest objection to changing the approach to restoration is that the commercial repair system easily handles 150 to 200 annual faults, and internet providers have spare capacity. But while the number of faults has remained flat, the number of cables and the lengths of new cables are increasing, complicating repairs. Clusters of near-simultaneous failures such as in Vietnam and Tonga resulted in restoration times stretching to months. Damage that may be deliberate is hard to distinguish from an accident. Such incidents have already occurred in the Baltic, where a Russia-linked tanker dragged its anchor across five cables in 2024, as well as near Taiwan, where a Chinese fishing boat and a Chinese freighter cut both cables to the Matsu Islands six days apart in 2023. And conflict has already complicated repairs in the Red Sea and in the Gulf. What has not happened yet is clustered failures, sabotage, and conflict all at once, a combination a wartime campaign could produce. A repair system sized for the rate of current accidents is not built to withstand an adversary that has learned to make attacks look like accidents.
A second objection is that low Earth orbit satellites can provide a backup. While they can keep some sites connected, they remain an access technology, not a backbone, and cannot replace the capacity of the cables they would be backing up. Satellite networks also depend on ground stations that are themselves connected by cable, and their signals can be jammed.
Adversaries are already building the means to cut cables deliberately. The Svalbard incident and news of China’s advanced deep-sea cable cutting device demonstrate that they are developing and testing this in areas that are challenging to repair. Russia’s Main Directorate of Deep-Sea Research operates at least eight nuclear-powered special-mission submarines, along with surface ships such as the Yantar, which carries submersibles capable of reaching depths of 6,000 meters (20,000 feet). Shadow fleet vessels can also be used to damage cables by dragging anchors, a common cause of cable faults. Strategic, heavily trafficked choke points in geopolitically contentious areas with dense concentrations of cables, such as the Suez Canal/Bab al-Mandeb Strait, Strait of Hormuz, and the South China Sea, provide an easy environment for multiple cuts that are hard to attribute. Additionally, an adversary could capitalize on naturally occurring faults, such as underwater landslides that may damage multiple systems at the same time, further stressing a repair system that is sized for “routine” cable faults.
How large would a campaign need to be? Consider Taiwan, where 15 international cables come ashore at four landing stations. When an earthquake off its east coast at the end of 2025 disrupted six of them, repairs took until May 2026. Taiwan sees seven to eight cable incidents a year, with dragged anchors the leading cause. Chinese vessels operating near Taiwan routinely switch off their location transponders. Cutting six cables deliberately, near shore where anchors can reach them, would require relatively few vessels. For example, a single tanker, the Eagle S, damaged five submarine cables in the Gulf of Finland in December 2024.
Repairs in a second theater would compete for the same few ships. The Red Sea cuts described above affected an estimated one quarter of the traffic between Asia, Europe, and the Middle East. Taiwan sits between the Yokohama and Southeast Asia repair zones. With an international cable repair taking about a month even in peacetime, a handful of ships facing nine faults across two theaters would need months to restore them.
Solutions
First, the Department of Homeland Security should formally designate submarine cables as a critical infrastructure subsector under the Communications Sector. Submarine cables span the Communications and Information Technology sectors, meaning no single agency owns and leads their restoration.
Second, Congress should designate a federal operational coordinator for submarine cable restoration with emergency authority to prioritize restoration when outages have national security and economic consequences. It should sit in the Department of Homeland Security, whose National Coordinating Center for Communications already coordinates the restoration of communications networks in emergencies. Restoration priorities should weigh the importance of military communications, financial networks, and cloud and data-center connectivity against the availability of redundant routes. That authority should include using Defense Production Act priority ratings for spares, which requires reauthorizing beyond Dec. 11, 2026.
Third, fund the Cable Security Fleet on a multiyear basis, expand it beyond two ships, and fund depots of spare cable parts. The current annual authorization for the fleet is $10 million for two ships. Doubling the fleet to four retained ships would raise the annual cost to around $20 million. Yet the Maritime Administration requested no funding for it in fiscal year 2027. With the recent decommissioning of the Navy cable ship Zeus, the United States should fund the procurement of a dedicated new cable repair ship, which costs between $52 million and $100 million to build and roughly $12 million to 14 million a year to run. Retained commercial ships are cheaper, but they might be on the other side of the world when called upon. A government-owned ship can be kept in a theater where it is needed.
This initiative should be paired with a government-funded and government-held inventory of spares free of Covered List components, pre-positioned at U.S. and allied ports. Using the European Union’s €60 million program as a benchmark, an inventory of spares covering the Pacific and Atlantic would cost at least as much. Taken together, a new ship, a doubled fleet, and a spares inventory would cost roughly $120 million to $170 million initially, and about $35 million a year to sustain. This is less than half of the €347 million (about $388 million) the European Union committed this year to cable projects under a single program. Additionally, the Maritime Administration should create pre-arranged procedures for activating war-risk insurance coverage for the fleet’s ships before a crisis. If commercial war-risk insurance is terminated, the agency could quickly offer government-backed coverage for both the ship and its personnel.
Fourth, establish a graduated framework for protecting repair operations in contested waters. Repair ships should remain under commercial control until there is a credible threat to them. At that point, the United States and its allies should escort ships working under government contracts, which are likely military objectives already, and provide security for commercial ships without placing them under convoy. A credible threat would include the withdrawal of war-risk cover, a warlike operations designation for an area, or an explicit adversary threat against cables. The federal coordinator proposed above should decide when that threshold has been met. The United States should also negotiate arrangements with allies, including command relationships, rules for escort, legal review, port access, and allied responsibilities, for repair operations in contested waters before a crisis.
Fifth, streamline the end-to-end process to build additional cable routes. The Federal Communications Commission’s 2025 order streamlined federal licensing, but constructing a new landing requires separate approvals from the Army Corps of Engineers, the National Oceanic and Atmospheric Administration if routes cross marine sanctuaries, state coastal agencies, and local zoning boards. The federal coordinator proposed above should oversee this process.
Commercial operators treat adversarial campaigns as a remote possibility. However, navies have long treated cables as targets. The U.S. Navy cut the telegraph cables serving Spanish Cuba in 1898, and during the Cold War, in an operation code-named Ivy Bells, the United States tapped a Soviet military cable in the Sea of Okhotsk. The Svalbard interception shows that America’s adversaries are rehearsing for covert undersea operations in waters where repairs are hardest. The United States has spent years deciding who may own, build, and operate subsea cables. It now needs to answer a simpler question: Who repairs the internet in wartime?
Anita Nikolich is a research scientist at the School of Information Sciences at the University of Illinois Urbana-Champaign.
Andrew Gallo is a senior network architect at the George Washington University.
Image: Lewis Clarke via Wikimedia Commons
