Protecting undersea internet cables is a tech nightmare

Strategic impact and who is hurt

• Several comments argue the U.S. mainland would feel limited direct impact from cuts because most user-facing services are hosted domestically; territories and globalized businesses would be hit harder.
• Europe is seen as more exposed due to reliance on transatlantic links and inter-EU undersea cables.
• China and, increasingly, Russia are described as more insulated because of domestic hosting and deliberate efforts to build semi-isolated or self-sufficient internets.
• Financial traffic (trillions in daily cross‑Atlantic transactions) is highlighted as a major vulnerability.

Redundancy, robustness, and economics

• Consensus: cables can’t realistically be “protected” end‑to‑end; resilience comes from many diverse routes.
• Laying extra cables is seen as cheaper and more effective than militarizing protection, though some question ongoing infrastructure and monitoring costs.
• Operational costs after deployment are said to be minimal; cables are “laid and forgotten” and even old, low‑capacity ones stay in service because retiring them is extra work.
• Some argue the Internet already routed around recent Baltic cuts with little user-visible impact, suggesting the threat is overhyped.

Accidental vs deliberate damage

• Ships accidentally cut cables roughly every few days; deliberate attacks are considered rare.
• In shallow seas like the Baltic, it’s hard to avoid crossing cables, which limits what redundancy can do.
• Dragging a large ship’s anchor is seen as a simple, deniable, and effective sabotage method.

Physical protection ideas

• Suggestions include: burying cables deeper, encasing them, or using plows/trenching tools along larger spans.
• Replies note burial is already used near shore and in shallow risky areas, but is expensive and makes repairs harder; deep‑water burial offers little extra benefit.
• Concepts like electrified “shock” cables or anchor-deflecting housings are dismissed as ineffective or too bulky.

Monitoring, policing, and deterrence

• Some advocate better tracking of ships near cable routes, faster naval response, stricter penalties for going “dark” (AIS off), and policies to reduce plausible deniability.
• Others counter that oceans are vast, attackers can use timed explosives or ROVs, and reaction time may not matter much.
• Piracy’s persistence is cited as evidence that even strong navies can’t fully police the seas.

Satellites and Starlink as backup

• Multiple comments stress that satellite capacity is orders of magnitude below modern fiber.
• Even Starlink’s entire RF capacity is said to struggle to match a single high‑end cable, and concentrating many satellites on one path is logistically limited.
• Satellite backup is viewed as useful for prioritized, emergency, or last‑mile traffic, but not a full substitute for transoceanic backbone capacity.
• NATO “future backup route” work is interpreted as proof‑of‑concept for such limited, prioritized roles, not a near‑term replacement for cables.

Geopolitics, treaties, and norms

• Some call for a new global treaty to protect undersea cables; others note an 1884 convention already exists but is largely untested.
• Skeptical voices argue that states already ignore international law when core interests are at stake; a new treaty would not stop clandestine operations.
• A counterview emphasizes that many international agreements do work in practice, that institutions like the UN Security Council and ICC have real (if imperfect) impact, and that norms and enforcement are incremental rather than all‑or‑nothing.

Sabotage, vandalism, and human behavior

• There is debate about how surprising it is that large shared infrastructures (from telegraph lines to modern cables) aren’t sabotaged more often “for fun.”
• Some point to abundant vulnerable infrastructure and occasional local vandalism/theft, but note that true large‑scale sabotage is empirically rare.
• Explanations include social cooperation norms, practical barriers (specialized equipment, boats, risk of getting caught), and the fact that most people are neither highly malicious nor very capable of targeted technical damage.