There is something about modern warfare. You can destroy a city without occupying it, paralyse a government without arresting a minister, and cripple a country's economy without firing a bullet at a bank. Increasingly, you can do all this by destroying the machines that keep the digital world alive. Welcome to the age when a warehouse full of computers may become as strategically important as an ammunition depot, and considerably more dangerous than its innocent-looking walls suggest.
Data centres are the new infrastructure of power. They host the systems supporting hospitals, financial transactions, emergency communications, public administration, logistics and the artificial intelligence increasingly used by military forces. They are the physical foundations of our supposedly weightless digital civilisation. The cloud, it turns out, has a concrete address, a power supply, cooling systems and a very real vulnerability to missiles.
And somebody, somewhere, is already calculating what happens if those facilities disappear. The problem is not simply that data centres can become military targets. Under international humanitarian law, civilian objects may lose their protection against attack when, by their nature, location, purpose or use, they make an effective contribution to military action and their destruction offers a definite military advantage. But the existence of military data inside a facility does not automatically turn the entire building into a legitimate target. Distinction, proportionality and precautions still apply. The difficulty is that these principles were not designed for an infrastructure where a military algorithm, a hospital database and millions of ordinary people's financial records may depend on the same physical machinery.
A server farm supporting battlefield intelligence might also host civilian services. A network used for military logistics might share infrastructure with commercial systems. An attack intended to blind a military command centre could simultaneously disrupt hospital appointments, banking transactions, emergency services and the supply chains delivering food and medicine. The missile strikes the building. The consequences travel everywhere.
This is the new arithmetic of collateral damage, and it is frighteningly easy to manipulate. Military planners can calculate the immediate advantage of destroying a digital facility. Calculating the cascading civilian consequences is harder, politically inconvenient and sometimes deliberately neglected. Governments love technological sophistication when it promises precision. They become remarkably vague when precision requires them to account for the thousands of people who might suffer because a supposedly surgical operation has severed the invisible nerves of civilian life.
Artificial intelligence makes the problem even more urgent. As armed forces integrate AI into intelligence analysis, targeting support, surveillance and operational planning, the infrastructure sustaining these systems becomes increasingly valuable. Yet greater military dependence does not justify treating every computer, cable or electricity substation as an enemy combatant. If anything, dependence should demand greater restraint.
Nor should we accept the comforting fiction that private technology companies are merely neutral suppliers standing outside the battlefield. When their infrastructure directly supports military operations, their facilities, contracts and operational decisions become part of a much larger security debate. That does not make every technology company a military actor, nor does it erase civilian protection. It means responsibility can no longer end at the corporate reception desk.
The Geneva Conventions do not need fashionable amendments simply because technology has changed. Their fundamental protections already apply to contemporary warfare. What is urgently needed is clearer international agreement on how those protections operate when military and civilian digital infrastructure overlap: how military objectives are identified, how foreseeable cascading harm is assessed, what precautions are feasible, and how accountability is established when the consequences extend far beyond the point of impact.
Governments should develop explicit operational rules, publish meaningful standards where security permits, and investigate attacks that may violate existing law. Technology companies must provide realistic assessments of civilian dependencies rather than hiding behind technical complexity. States should also reduce dangerous concentrations of essential services, strengthen backups and build resilient alternatives before a crisis begins. Resilience is not glamorous. Neither is explaining to a patient why the hospital's systems have vanished.
And let us abandon another convenient illusion: that a digital war is somehow cleaner than a conventional one. A server does not bleed, but the people depending on it certainly can. A destroyed financial platform may leave families unable to pay for necessities. A disabled hospital system may delay treatment. A broken communications network may separate civilians from evacuation information. The absence of visible bodies in the server room does not make the consequences bloodless.
The next generation of war crimes will not necessarily look like the last. Some will still involve the familiar horrors of massacres, torture and deliberate attacks on civilians. Others may emerge from decisions made in command centres thousands of kilometres away, where an apparently technical calculation conceals a devastating human cost.
We should not wait for the first catastrophic attack on a shared civilian digital backbone to discover that our laws, institutions and infrastructure were unprepared. The cloud is not above the battlefield. It is inside it. And when the servers go dark, the people left in the darkness will not be algorithms.
No comments:
Post a Comment