Researchers at China's military have warned that the spread of low-cost unmanned underwater vehicles (UUVs) is raising the risk of attacks on undersea communications cables, the South China Morning Post (SCMP) reported on August 5, 2026. The warning was aimed at the first Pillar II flagship project announced in May by AUKUS, the US-UK-Australia security framework. The three countries will develop common UUV payload systems covering everything from protecting critical undersea infrastructure to reconnaissance and attack, with deployment beginning in 2027. The "99%" associated with undersea cables is not a loss rate in an attack. It is the share of the world's international data that the cables carry. The competition is shifting away from a weapon that could take down the entire internet in one blow and toward the ability to routinely operate networks of underwater robots that serve both defense and offense.
Starting in 2027: AUKUS's first Pillar II flagship project
On May 30, 2026, the three AUKUS nations announced a plan to jointly develop payloads and underlying systems for UUVs. While Pillar I deals with introducing nuclear-powered submarines to Australia, Pillar II is a framework for the three countries to develop advanced military technologies together. This project is the first under Pillar II to be labeled a "signature project."
Deployment will proceed in two stages. First, each country will develop payloads that deliver different effects, designed so they can be swapped onto the UUVs of any of the three nations. Then the countries will jointly develop and produce next-generation payloads and the technologies that support their operation. The UK government has said initial capability is expected to become operational in 2027, and the Royal Navy will also be able to integrate US- and Australian-made payloads.
What is being standardized is less the vehicle itself than the functions it carries and the mechanisms that link them. Australia's Department of Defence cited sensors, navigation, and strike capabilities as examples of payloads. The underlying systems will support information sharing among UUVs and coordination with crewed ships and submarines. Swapping payloads by mission and running them on the same control system makes it easier to speed up deployment than preparing separate dedicated vehicles for each sea area.
Common payloads that span infrastructure protection and attack
The missions listed by Australia's Department of Defence range widely, from protecting undersea infrastructure to surveillance, reconnaissance, and attack. The same project also includes logistics and mine countermeasures, anti-submarine and anti-surface warfare, and electronic warfare. A sensor network that patrols along cables to look for anomalies can also serve as a military network that continuously tracks conditions underwater. AUKUS's own published materials give a military rationale to the warning SCMP relayed.
UUVs are envisioned to work alongside crewed ships and submarines. However, AUKUS has not disclosed the number of vehicles, payload performance, operating areas, or total project cost. Nor are there any confirmed test results showing that low-cost vehicles can actually cut cables. The only change that can be confirmed at this point is that the three countries have moved toward a 2027 deployment of common payloads for multipurpose UUVs, including for undersea infrastructure protection.
The same technology can be used both to detect attacks and to prepare them. In 2024, the three countries remotely operated unmanned systems off Australia from the US and systems off the US from Australia. In 2025, Japan joined, and acoustic communications were used to give maritime tasking to a large UUV. For this project too, Australia's Department of Defence says common standards, a shared operating concept, and a shared control system will improve interoperability. What matters is less the performance of individual drones than whether multiple vehicles and crewed forces can be run as a single unit.
The network carrying 99%: 70–80% of faults are accidents
According to the International Telecommunication Union (ITU), undersea communications cables carry more than 99% of the world's international data traffic. The International Cable Protection Committee (ICPC) puts the global undersea network at roughly 500 systems with a total length of about 1.8 million km. Whether an individual fault leads directly to a service outage depends on whether traffic can be rerouted to other cables or terrestrial links.
Meanwhile, of the 150 to 200 cable faults that occur each year, 70–80% are caused by accidental human activity such as fishing and ship anchors. The ICPC says it holds no data on state-sponsored sabotage and takes the view that there have been no verified incidents since World War II. Because the committee is not an agency that investigates the intent behind incidents, this is not an explanation ruling out the possibility of sabotage. Even so, to reduce everyday outage risk, measures to prevent damage from fishing and anchoring would work before military UUV patrols.
The phrase "99% is at risk" blurs these two figures together. The 99% is the share of data carried by undersea cables as a whole; it does not mean that a single inexpensive UUV could cut off that same share. Carriers design their networks on the assumption that they can reroute traffic over multiple paths and repair fault locations. The degree of danger depends less on the number of cables than on how concentrated routes and landing stations are in a particular region.
The waters to be monitored are not uniform either. According to the ICPC, faults are most frequent at depths of less than 300 m, where cables are more exposed to ships and fishing activity. Searching broadly across the deep seabed of the open ocean and frequently monitoring vessels and fishing gear approaching landing points call for different vehicles and sensors. Since AUKUS's published materials do not state depth, detection range, or recharging methods, it is not yet possible to judge which risk the plan will reduce first.
Rerouting and repair time matter before surveillance networks
On July 10, 2026, an international advisory body established by the ITU and ICPC approved its final report on undersea cable resilience. The challenges it identified are exposure to physical risks, lengthening repair times, the geographic concentration of infrastructure, and dependence on a small number of cables. Countermeasures include public-private information sharing, shorter permitting times, and risk monitoring, alongside route diversification and redundancy.
UUVs can shorten the time needed to find anomalies on the seabed and identify approaching objects. But they do not create the capacity to shift severed communications onto other routes, nor the permits that repair ships need to head to the site. The surveillance handled by militaries and the rerouting and repair handled by private operators need to be developed separately.
What to check in 2027 is less how many vehicles AUKUS fields than at what depths and endurance the initial payloads can detect anomalies, and whether they can be exchanged among the three countries' existing UUVs. Once those specifications are clear, it will be possible to tell whether undersea cable defense is becoming a practical surveillance capability, or whether building a military network that includes attack missions is moving ahead first.