We Will Save the Birds From AI — But Who Will Save Humans From the Bomb?

Author: Rodgers Mangwela

---

There is something deeply uncomfortable about the way modern civilisation chooses its battles.

We can become furious when an artificial intelligence data centre threatens electricity supplies, consumes water or disrupts birds and wildlife. Musicians can protest when AI-generated songs imitate their voices or compete with human creativity. Environmental campaigners can demand that technology companies disclose their carbon footprint.

And these concerns are real.

But somewhere else, behind fences, security checkpoints and national flags, factories are producing missiles, drones, artillery systems and other weapons whose intended purpose is not merely to disrupt a bird's habitat or replace a musician.

They are designed, ultimately, to destroy targets.

The question is not whether AI's environmental and social consequences deserve scrutiny. They do. The more uncomfortable question is this: if protecting life is the principle, why does the principle sometimes appear to stop at the factory gate?

AI's growing appetite for energy and water

The environmental concerns surrounding artificial intelligence are not imaginary.

The International Energy Agency estimates that data centres consumed about 415 terawatt-hours of electricity in 2024, roughly 1.5% of global electricity consumption. The agency expects global data-centre electricity demand to continue rising sharply as AI expands. Its 2025 analysis projected that electricity generation serving data centres could exceed 1,000 TWh by 2030 in its base case.

The environmental footprint goes beyond electricity.

Data centres may require substantial water for cooling, although the amount varies considerably depending on technology, location and climate. The UN Environment Programme has warned that growing data-centre demand can increase pressure on both energy and water resources.

There are also concerns about minerals, electronic waste and the manufacturing of the chips and servers needed to operate AI systems.

These are legitimate reasons for regulation, transparency and better engineering.

But there is an important distinction: AI itself is not a weapon, and its environmental footprint is not equivalent to the direct human toll of armed conflict.

That distinction makes the comparison more uncomfortable, not less.

When technology threatens the creative worker

For artists, the debate becomes personal.

A songwriter can spend weeks writing a song, a singer can spend years developing a voice and a producer can build a career around a distinctive sound. Generative AI can then produce music within seconds.

The question is not simply whether AI can make music. It is whether the economic system surrounding AI will fairly compensate the humans whose work helped create the cultural material on which these systems are trained.

A global study commissioned by the International Confederation of Societies of Authors and Composers, or CISAC, estimated that under an unchanged regulatory framework, 24% of music creators' revenues could be at risk by 2028. The study also projected that AI-generated music could become a significant part of the commercial music market.

That is a serious concern for musicians.

It is particularly relevant in Africa, where many independent artists already struggle with copyright enforcement, distribution, streaming income and access to professional infrastructure.

For publications such as KASONDE24, the wider lesson is bigger than music: technological change should be examined not only through the excitement of what machines can create, but through what happens to the people whose livelihoods are caught in the transition.

Yet even here, there is another question.

If society is capable of asking whether an algorithm has harmed a songwriter's livelihood, why should it be incapable of asking whether an industry has contributed to the destruction of a human community?

The industry that humanity calls necessary

Weapons are different because governments generally justify military production through national security.

Countries argue that weapons are necessary for deterrence, defence and sovereignty. A state facing a military threat may regard the ability to produce weapons as essential to protecting its population.

That argument cannot simply be dismissed.

But necessity does not make consequences disappear.

According to the Stockholm International Peace Research Institute, global military expenditure reached an estimated $2.887 trillion in 2025, marking the 11th consecutive year of growth. Military spending represented about 2.5% of global GDP.

That is an extraordinary amount of economic activity connected to national security.

Behind the statistics are factories, engineers, scientists, supply chains, contractors, governments and workers.

There are also cities.

There are families.

There are people who wake up to the sound of explosions rather than the sound of birds.

The human cost hidden behind the numbers

The UN's data on armed conflict provides a sobering counterpoint to discussions about technological risk.

In Ukraine alone, UN human rights monitors verified 1,839 civilians killed and 10,638 injured during the first seven months of 2026. Long-range missiles and drones were among the leading causes of civilian casualties.

Globally, the UN reported that conflict-related civilian deaths fell in 2025 from the exceptionally high level recorded in 2024, but still reached 37,163 recorded civilian deaths in conflicts where data was available. Heavy weapons and explosive munitions were identified as the leading cause of death among the categories measured.

These figures should not be used to suggest that every weapon manufacturer, soldier or government intends to kill civilians. International humanitarian law distinguishes between legitimate military objectives and protected civilians, and states routinely describe their weapons programmes in terms of defence.

But the existence of civilian casualties reminds us of something technology debates sometimes obscure:

Every technological system eventually operates inside a human society.

Why do we react differently?

Perhaps the difference lies partly in distance.

An AI data centre can be photographed. Its electricity consumption can be measured. Its cooling systems can be inspected. A campaign against an environmentally damaging facility can produce a clear image of machines, cables, cooling towers and disappearing water.

War is different.

Its consequences may occur thousands of kilometres away from the people buying the weapons.

A missile factory can employ thousands of people. Its products may generate billions of dollars in contracts. The destruction happens elsewhere, often after the original transaction has disappeared behind layers of government procurement, military planning and geopolitical justification.

This does not mean environmental campaigns are wrong.

It means that human beings may be more comfortable confronting harms when the chain between production and consequence is visible.

Follow the money, but follow the purpose too

Both AI and the military-industrial economy demonstrate the enormous power of commercial and strategic incentives.

Technology companies invest billions because they expect AI to transform industries and generate revenue. Governments invest in defence because they believe military capability provides security in an uncertain world.

Neither motivation automatically makes an industry immoral.

But neither should make an industry immune from scrutiny.

The same society that demands transparency from technology companies about electricity, water and copyright can reasonably demand transparency about military spending, weapons exports and civilian protection.

The standards may not be identical, because the risks are different.

But the underlying question can be similar: What is this technology doing to human beings, and who bears the cost?

Perhaps the real debate is about what we value

There is a danger in turning this into a simple argument against AI.

AI can help scientists analyse climate data, improve energy systems, assist medical research and expand access to information. The IEA also notes that AI could help optimise parts of the energy system, meaning its environmental impact cannot be understood only through its electricity consumption.

Likewise, defence technologies can serve legitimate protective purposes. A military capability can be presented as a deterrent precisely because governments hope it prevents war.

The deeper issue is therefore not whether one technology is "good" and another is "bad".

It is whether humanity applies its ethical scrutiny consistently.

If we ask how many litres of water a data centre consumes, perhaps we should also ask how many resources are consumed by an arms race.

If we worry about AI replacing musicians, perhaps we should also ask what happens to musicians when war destroys their homes, studios and audiences.

If we care about protecting birds from technological disruption, perhaps we should also care about protecting the human beings living beneath the flight paths of missiles and drones.

The future will not be decided simply by whether machines become more powerful.

It will be decided by the values that guide the people who build, regulate, buy and use them.

And perhaps that is the uncomfortable question worth carrying beyond the headlines:

If protecting life is truly the principle, how far does that principle extend ? From the forest, to the data centre, to the recording studio, and finally to the factory producing the bomb?