Submarine Cable Landing Station Risk Profile in India 2026: Vizag Gateway, Cable Cut Exposure and the Connectivity Asset Programme

For three decades, India's international internet has come ashore at two clusters: Mumbai on the west coast, carrying the majority of inbound traffic, and Chennai on the east. Visakhapatnam is now becoming the third gateway, and the reason is the AI build-out, not legacy telecom.

In late 2025 Airtel announced it would construct a new cable landing station (CLS) at Vizag to host Google's subsea systems under the America-India Connect initiative, part of Google's stated multi-billion-dollar Indian infrastructure programme running through the end of the decade. Alongside it sits a gigawatt-scale AI data-centre campus being developed with AdaniConneX. Meta's Project Waterworth, a roughly 50,000 km, 24 fibre-pair system spanning five continents, has selected Mumbai and Vizag as its two Indian landing points, with the system expected to come live around the end of the decade.

The practical effect is that a single coastal site near Vizag will concentrate three asset classes that insurers normally underwrite in separate silos: the landing station building and its powered terminal equipment (property and engineering), the nearshore and beach-manhole cable segment (marine), and the dependent data-centre load that monetises the capacity (business interruption and contingent BI).

There is a regulatory layer the broker cannot ignore. A landing station is licensed infrastructure: the cable system needs Indian landing-party clearances, and operating a CLS sits under telecom licensing oversight, with security-agency conditions attached to who may land a foreign cable and how it is monitored. Those licence conditions can drive contractual liabilities, indemnities owed to consortium partners and uptime commitments that flow straight into the business-interruption exposure the programme has to answer. A broker who reads only the property schedule and never the landing-party agreement will misjudge both the sums at risk and where liability actually sits.

For brokers, the live question through 2026 and 2027 is who carries which segment, and whether the wordings stitched together actually close the seam. India's broader vulnerability, highlighted after the September 2025 anchor-drag cuts off the subcontinent, makes that seam commercially material rather than theoretical.

Before placing anything, map the loss scenarios a Vizag-style CLS actually faces. They do not behave like a factory fire.

Nearshore cable cut

The dominant peril worldwide is mechanical: ship anchors dragged in shallow approach waters, and fishing gear. India recorded anchor-drag cuts in September 2025. The nearshore segment, the few kilometres between the beach manhole and deeper water, is where most cuts happen and where repair is slowest because shallow-water work needs specialist shallow-draught support.

Repair delay

India has no domestic cable repair ship. Operators contract consortium vessels based out of Dubai and Singapore under multi-year agreements. A vessel may be days or weeks away on another fault when yours occurs. Repair lead time, not the splice itself, drives the BI quantum.

Landing station property loss

Fire, flood or cyclone at the CLS building. The east coast is cyclone-exposed; Andhra Pradesh sees periodic severe systems. Powered terminal equipment (line terminating equipment, power feed equipment) is high-value and replacement-lead-time sensitive.

Power and cooling failure

Like any data hall, a CLS depends on continuous power. A protracted outage that trips terminal equipment is a non-damage interruption that many property wordings exclude.

Contingent dependency

The gigawatt data-centre campus monetises the landed capacity. A cable cut that does not touch the data centre at all can still strand its connectivity, a contingent BI loss with no material damage on the insured's own premises.

Each scenario lands on a different policy. The placement either coordinates them or leaves the insured arguing causation between insurers.

The single most common error in placing CLS risk is treating the whole asset as property. The cable in the water is a marine asset, and standard fire or industrial all-risk wordings either exclude it or fall silent.

Wet-segment cable is conventionally insured under a specialist marine cable policy, drawing on hull and marine cargo principles rather than fire-policy logic. The exposures it must answer are particular:

• Repair cost including vessel mobilisation, which can dwarf the cost of the cable itself. Mobilising a consortium ship from Singapore to an Indian fault is the headline number.
• Cable cut by third party, where subrogation against the offending vessel's owner is theoretically available but practically hard, because the ship is often unidentified, foreign-flagged or judgment-proof.
• General average and salvage principles where a repair operation interacts with a casualty, familiar to anyone who has handled marine hull general average matters.

The boundary problem is the beach manhole. Landward of it, the cable is arguably terrestrial property; seaward, it is marine. If the property and marine policies define that handover differently, a cut in the surf zone can fall into a gap both insurers decline. The fix is mechanical, not clever: pick a single physical reference point, name it in both wordings, and state in writing that the marine policy responds up to and the property policy responds from that exact point. A latitude and longitude or a named manhole chamber is worth more here than any amount of broad coverage language, because the dispute, when it comes, is about where the loss happened, not whether the peril was insured.

Brokers should treat the policy wording for the cable segment as the controlling document and build the property cover around it, not the other way round. The marine market understands cable; the fire market generally does not.

The CLS building and its powered plant sit closer to home for most Indian insurers, but the values and lead times are atypical.

The terminal equipment, line terminating equipment, power feed equipment and the optical gear that lights the fibre, is specialised, often single-sourced and slow to replace. That makes the reinstatement-value basis and the indemnity period the two numbers that decide whether a claim leaves the insured whole. A standard 12-month indemnity period is frequently too short when replacement of bespoke terminal equipment, sourced from a handful of global vendors, runs well beyond a year from order to commissioning.

Key placement points:

• Sum insured on a reinstatement basis, with terminal equipment valued at current replacement plus installation and commissioning, not depreciated book value. Watch the average clause carefully; underinsurance on high-value terminal racks bites hard.
• Machinery breakdown and electronic equipment cover for the powered plant, the same engineering logic discussed in our data-centre operator risk profile.
• Catastrophe sub-limits appropriate to an east-coast cyclone zone. Andhra Pradesh's coast takes periodic severe systems; flood and storm surge at a beach-adjacent site are live perils, not boilerplate.
• Contractors' or erection all-risks during construction, with advance loss of profits if the CLS commissioning gates a dependent revenue stream, the structure we set out in our single-project EAR for mega infrastructure note.

The material damage cover and the BI cover must share one definition of the insured property and one indemnity period. Where they diverge, the BI claim stalls on whether the trigger event was a covered material-damage event at all.

BI is where a CLS programme earns or loses its money, and where the contingent layer is most often missing.

Direct BI responds when a material-damage event at the insured CLS (fire, flood, equipment failure) suspends operations. The driver of quantum is the indemnity period set against realistic repair and replacement lead times, including the wait for a foreign repair vessel. A business-interruption calculation built on a 12-month period will understate a loss where terminal equipment or a nearshore splice takes longer to restore. Our BI claim quantification guide sets out the gross-profit mechanics that apply.

The harder layer is contingent. The gigawatt data-centre campus does not own the cable, yet its revenue depends on landed capacity. A nearshore cut that never touches the data hall still strands its international connectivity. That is a consequential-loss with no material damage on the insured's own premises, exactly the trigger that standard BI excludes and that contingent business-interruption cover is built to answer.

Two refinements matter for Vizag specifically:

1. Non-damage business interruption for a power or cooling outage that trips terminal equipment without physical damage. Standard wordings exclude it; the same gap we flagged in utility and telecom outage cover applies here.
2. Denial of access where a port closure, naval restriction or coastal incident blocks the CLS or the repair operation without damaging it, the scenario set out in our denial-of-access BI work.

The practitioner instruction is blunt: never place the CLS property and the dependent data-centre BI as if they were one risk on one schedule. Map the dependency explicitly, name the contingent property, and set the indemnity period to the slowest realistic restoration path.

A landing station is critical national connectivity infrastructure, which changes its threat profile beyond the physical.

Cyber exposure is twofold. The terminal equipment and network management systems are attack surfaces, and a successful intrusion can interrupt traffic without any physical damage, a cyber business-interruption trigger that property cover will not touch. The placement needs an affirmative cyber policy that responds to operational technology interruption, not just data-breach liability.

The sharper issue for 2026 is the war and political-violence boundary. India's subsea routes run through contested waters, and the Red Sea corridor has been repeatedly flagged as a threat to the country's digital lifeline. Sabotage, suspected state-linked anchor-dragging and conflict-zone exposure sit at the edge of, or outside, standard marine and property wordings.

• War and terrorism: most property and marine wordings carry war exclusions. A CLS placement should resolve explicitly whether deliberate cable cutting in contested waters is covered, excluded or carved into a separate terrorism-insurance or marine war facility.
• State-linked sabotage: attribution is hard, and the line between a hostile act and an accidental anchor drag is exactly where claims are contested.

India's security agencies have publicly flagged subsea cable resilience, including reported plans to retrofit naval vessels for repair support. Underwriting that ignores the security dimension of a landing station is underwriting half the risk.

The broker's job is to make the war and cyber boundaries explicit at placement, in writing, rather than discovering them in a contested claim after a cut whose cause is genuinely ambiguous.

Pull the threads into a placement discipline. A Vizag-class CLS is a multi-policy programme, and the value the broker adds is in coordinating the seams, not in any single section.

1. Segment the asset on a single schedule of definitions. Beach manhole, nearshore cable, deep-water cable, terminal equipment, building, dependent data centre. Every policy in the tower must reference the same boundaries.
2. Place the marine cable cover first and build property around it, so the beach-manhole handover is owned by one wording, not contested between two.
3. Set the indemnity period to the slowest realistic restoration, factoring foreign repair-vessel mobilisation, not a default 12 months. Document the assumption.
4. Name the contingent property explicitly. The dependent data centre must be identified by name in the contingent BI section, with its own dependency limit.
5. Resolve war, sabotage and cyber boundaries in writing before binding, including whether deliberate cable cutting is in or out.
6. Stress-test underinsurance on terminal equipment values against the average clause; reinstatement valuation, not book value.
7. Confirm subrogation strategy for third-party vessel cuts, accepting recovery is often impractical against unidentified foreign ships.

The capacity picture matters too. This is specialised risk; Indian insurers will lean on facultative reinsurance and London or Singapore marine markets for the cable and war layers. GIC Re's appetite and IRDAI's treatment of large project placements shape how much sits onshore.

For a corporate risk manager, the single test is this: if a ship drags an anchor across the nearshore segment tomorrow, can you name the policy that pays, the indemnity period that applies, and the dependency limit for the data centre, without a coverage argument? If not, the programme is not finished.