Everyone's talking about undersea cable sabotage. Half of what you've heard is wrong.

If you’ve spent any time on geopolitical Twitter or defence think-tank feeds lately, you’ve seen the warning: “Watch the undersea cables.” The story goes that Russia and China are systematically sabotaging the internet’s physical backbone. Nine Baltic Sea incidents since 2022. Forty-four cables damaged. A new Chinese submarine built specifically to cut cables. Australia supposedly has only three links to the outside world.

It’s a gripping narrative. Parts of it are true. Parts of it are not. I spent the last few days running every major claim through primary sources. Here’s what held up and what fell apart.

What’s actually happening

The Baltic Sea has become ground zero for undersea infrastructure damage. Since September 2022, there have been at least seven confirmed incidents where cables or pipelines on the Baltic seabed were severed. This is verified through Reuters, AP, CNN, and the gCaptain maritime timeline.

The pattern is remarkably consistent each time. A cargo ship - usually flagged to a third country, often with opaque ownership, frequently departed from a Russian port - drags its anchor across the seabed. Cables get cut. Authorities seize the ship. Investigators find anchor drag marks stretching for kilometres. And then the courts can’t prove it was deliberate.

The most recent case was New Year’s Eve 2025. Finnish telecom company Elisa noticed a disruption on its Helsinki-to-Tallinn cable at 4.53am. Finnish Border Guard dropped soldiers from a helicopter onto the deck of a cargo ship called the Fitburg, which had just left St Petersburg. Anchor drag marks were found on the seabed. Two crew members were arrested.

The same script played out in December 2024 with a tanker called Eagle S, which dragged an 11,000 kilogram anchor for 90 kilometres across the Gulf of Finland, cutting a power cable and four telecom lines. Finland seized the ship and confirmed it was part of Russia’s “shadow fleet” - vessels operating with opaque registrations to evade oil sanctions.

And in February 2023, both cables connecting Taiwan to its Matsu Islands were severed by Chinese vessels. Fourteen thousand residents lost internet for nearly two months. Taiwan blamed China. Matsu’s cables had already been damaged 30 times since 2017.

None of this is disputed. The incidents happened. The ships were identified. The damage is documented.

What’s not true

Here’s where the narrative breaks down.

Claim: “80 per cent of US military communications travel via undersea cables.”

This statistic has been repeated hundreds of times. I could not find a primary source for it. Aaron Bateman, the George Washington University professor most commonly cited in connection with this figure, wrote a detailed essay on cable vulnerability for Engelsberg Ideas. It does not contain this statistic. No Department of Defense source was found. The claim appears to be an extrapolation from the “97 per cent of intercontinental data” figure, applied to military traffic without evidence.

Claim: “Fifteen undersea cables were severed in the Red Sea in September 2025.”

This one conflates two different numbers. Three cables were confirmed cut near Jeddah - SEA-ME-WE 4, IMEWE, and FALCON. Fifteen is the total number of cables passing through the Bab el-Mandeb Strait, not the number severed. The International Cable Protection Committee attributed the cuts to commercial shipping activity, likely an anchor drag. Yes, it caused real disruption - up to 30 per cent latency increase between India and Europe. But “15 cables severed” is a factual error that has been repeated across dozens of articles.

Claim: “Australia has only three undersea cable connections and is extremely vulnerable.”

This is just wrong. Australia has 12 to 14 international submarine cables with a total capacity exceeding 350 terabits per second. The three cables usually named - Indigo, JGA-S, and Oman Australia Cable - are real but represent only a fraction. Missing from the typical list: Southern Cross (92 Tbps to the US), Southern Cross NEXT (72 Tbps), Hawaiki (30 Tbps to the US), the Australia-Singapore Cable (60 Tbps), and several others. The Australian Communications and Media Authority maintains an official map, updated January 2026, that shows the full picture.

Australia is less connected than northern hemisphere hubs, and its cables are geographically concentrated at Sydney and Perth, creating genuine chokepoint risk. But “three cables” drastically understates the reality.

Claim: “88 per cent of advanced semiconductor trade passes through the Taiwan Strait.”

No primary source exists for this figure. After extensive searching across the US International Trade Commission, CSIS, Bloomberg, and the US-Taiwan Business Council, I could not find anyone who has published this specific statistic. It appears to be a conflation with a real and well-sourced figure: Taiwan manufactures approximately 92 per cent of the world’s most advanced semiconductors (USITC, February 2024). Someone appears to have confused “made in Taiwan” with “shipped through the strait.”

The Finnish intelligence problem

Here’s the claim that does the most damage to the “systematic sabotage” narrative, and it comes from a NATO member’s own intelligence service.

Finland’s domestic intelligence agency, SUPO, published a statement in March 2026 that said: “Cable damages have been a common occurrence in the Baltic Sea for decades. Ship anchors have been regularly damaging cables in the Baltic Sea throughout the 21st century, even though the incidents have not made the news. Statistically, the number of cases of cable damage in recent years is quite typical for the Baltic Sea.”

That’s not a Russian propaganda outlet. That’s Finnish intelligence. A country that shares a 1,340-kilometre border with Russia. A country that just seized two ships. A country that is arguably more hawkish on Russia than any other NATO member.

SUPO also noted that “the incidents of ships dragging their anchors in recent years have damaged not only Western undersea infrastructure but also Russia’s own infrastructure.” If Russia were deliberately ordering ships to drag anchors, why would it damage its own cables?

The International Cable Protection Committee, the primary global industry body for cable protection, published a position paper in February 2025 that said media coverage “can lead to the speculation that such incidents are uncommon or suspicious - however such incidents can be attributed to poor seamanship practices, equipment failure or improper stowage of anchors.”

And Recorded Future - the threat intelligence company whose “44 cable damages” statistic is cited in virtually every article on this topic - admitted in its own July 2025 report that “the increase may be partially attributable to increased reporting on subsea cable faults, as opposed to strictly reflecting an increase in incidents.”

That’s the company whose business model depends on threat amplification, acknowledging that the apparent spike might be a reporting artifact.

The other side

Even accidental cuts create genuine vulnerability. A cable severed by a dragged anchor takes days or weeks to repair, regardless of intent. The Baltic Sea is shallow, narrow, and crowded. More cables plus more ships equals more incidents, even without any conspiracy. But whether an incident is deliberate or accidental matters less than the fact that the infrastructure has no redundancy in many corridors. One cut cable can isolate a region.

The lack of successful prosecutions deserves scrutiny. Maritime law creates enormous barriers to proving intent. The Helsinki court dismissed the Eagle S case on jurisdiction, not on the merits. The ship’s flag state, the Cook Islands, will never prosecute. That legal gap means zero convictions does not equal zero guilt. It may simply mean the legal architecture is not built for this kind of case.

Russia’s shadow fleet of ageing, poorly maintained tankers creates plausible deniability. A ship with failing equipment, an inexperienced crew, and opaque ownership can damage a cable, and the structural conditions make it nearly impossible to distinguish negligence from sabotage. That ambiguity may be a feature, not a bug.

What is genuinely concerning

Despite all the corrections above, there are real reasons to pay attention.

China publicly disclosed a deep-sea cable-cutting device in February 2025, developed by the China Ship Scientific Research Centre. It can operate at 4,000 metres depth and sever armoured cables. It was published in a Chinese academic journal. This is not Western speculation - China itself disclosed the capability.

Russia’s GUGI (Main Directorate for Deep Sea Research) operates sophisticated submarines including the Losharik, capable of diving to 6,000 metres, specifically designed for seabed operations. Russian naval doctrine, while not explicitly calling for cable attacks, emphasises seabed control. Dmitry Medvedev has publicly hinted at targeting cables as retaliation for the Nord Stream pipeline destruction.

The Taiwan Matsu Islands incident in 2023 was almost certainly not an accident. Chinese vessels cut both cables to a Taiwanese territory, in a pattern consistent with previous Chinese sand-dredging intrusions. It was widely interpreted as a rehearsal for cutting Taiwan off from the outside world during a military scenario.

NATO launched Baltic Sentry in January 2025 - a standing year-round mission with frigates, maritime patrol aircraft, and naval drones - specifically to protect undersea infrastructure. This is not theatre. NATO does not maintain standing missions for imaginary threats.

The honest picture

The truth is messier than the viral version.

Most Baltic cable incidents are probably accidental. The Baltic is shallow, narrow, and crowded with shipping. Russia’s shadow fleet consists of ageing tankers with poor maintenance. Anchor failures happen. Ship traffic has increased. Cable density has doubled in recent years. More cables plus more ships plus shallow water equals more incidents, even without any conspiracy.

Some incidents may involve deliberate or semi-deliberate acts by individual vessel operators. But proving that the Russian state ordered a specific crew to drag an anchor over a specific cable has proven impossible in every attempt.

What is genuinely real and genuinely concerning is the preparation for wartime cable disruption. Russia and China are building the capability to cut cables during a conflict. China has disclosed a purpose-built tool. Russia has the submarines. Taiwan has already been practised on. The peacetime incidents might be mostly accidents, but the wartime preparation is real.

The cables are worth watching. Just watch them with open eyes.

The way forward

NATO’s Baltic Sentry mission is the right model. A standing naval presence with frigates, patrol aircraft, and drones in high-risk corridors makes deliberate acts riskier and accidental ones easier to investigate. That framework should be expanded to other vulnerable regions.

Australia should map and publicise its full cable infrastructure. The ACMA already maintains an official cable map, but most Australians - and most policymakers - could not name a single cable beyond the three that get repeated online. Transparency builds public support for investment in resilience.

Cable repair ships are a critical gap. There are only a handful of specialised cable repair vessels operating globally. None are based in the Asia-Pacific south of Singapore. A cable cut off Sydney could wait weeks for a repair ship to arrive. Australia and its regional partners should fund a rapid-response repair capability based in the southern hemisphere.

The international legal framework for investigating maritime cable damage is inadequate. UNCLOS gives jurisdiction to the flag state, which in practice means shadow fleet ships answer to flag-of-convenience registries with no capacity or incentive to investigate. The international community needs a legal mechanism - perhaps through the International Maritime Organization - that allows coastal states to investigate and prosecute cable damage in their exclusive economic zones, regardless of flag state consent.

Landing point diversification matters. Most of Australia’s international cables come ashore at two locations: Sydney and Perth. A single natural disaster or targeted event at either point could sever a large share of the country’s international bandwidth. Future cables should be routed to additional landing points - Melbourne, Brisbane, or Darwin - to reduce geographic concentration.

Sources

Primary sources for this article include: Reuters (multiple reports Oct 2025-Feb 2026), AP News, BBC, CNN, NATO official announcement (Jan 14 2025), Finnish Border Guard statements via Yle, International Cable Protection Committee position paper (Feb 2025), Recorded Future Insikt Group report (Jul 17 2025), Finnish intelligence SUPO statement (Mar 2026), South China Morning Post (Mar 22 2025), US International Trade Commission briefing (Feb 2024), CSIS Crossroads of Commerce report (Aug 2024), Australian Communications and Media Authority official cable map (Jan 2026), Bulletin of the Atomic Scientists “Seabed Zero” (Feb 13 2026), gCaptain Baltic timeline (Jan 2026).