Even if we grant, for the sake of argument, that global net zero is desirable in principle, there remains an increasingly uncomfortable question: is it actually going to happen?
The latest global energy figures make that question difficult to avoid. As Macrobusiness notes, citing the 2026 Statistical Review of World Energy, global consumption of oil, natural gas and coal has reached record levels. Fossil fuels still provide the overwhelming majority of the world's energy, while carbon emissions associated with fossil-fuel consumption have also reached record levels. Australia, despite years of political commitment to renewable energy and emissions reduction, remains overwhelmingly dependent upon fossil fuels when total energy consumption rather than electricity generation alone is considered.
That is the reality from which net zero must begin. We are not attempting to decarbonise a stationary global energy system. We are attempting to decarbonise one whose appetite for energy continues to increase.
This distinction is crucial. If global energy consumption remained constant, replacing fossil-fuel infrastructure with nuclear, solar, wind, hydroelectricity and other low-emissions technologies would already constitute one of the largest engineering projects in human history. But demand is not remaining constant. The world is entering what the International Energy Agency itself now calls the "Age of Electricity."
The IEA expects global electricity demand to grow by an average of 3.6 per cent annually between 2026 and 2030. That sounds modest until one considers the enormous existing scale of the electricity system. Global consumption is forecast to rise from around 28,200 terawatt-hours in 2025 to 33,600 TWh in 2030. In other words, in only five years humanity must find roughly another 5,400 TWh of annual electricity generation merely to satisfy additional demand.
China is central to this story. The IEA forecasts Chinese electricity consumption increasing at an average 4.9 per cent annually through 2030. In absolute terms, China alone is expected to add approximately 2,600 TWh of annual electricity demand over five years. That increase is roughly equivalent to the entire present electricity consumption of the European Union.
Then comes India. Its electricity demand is forecast to grow by approximately 6.4 per cent annually through 2030, adding more than 570 TWh of annual consumption. Rising incomes, industrialisation, urbanisation and air-conditioning are driving demand. This is hardly surprising. Hundreds of millions of people reasonably aspire to the energy-intensive living standards that Western populations have enjoyed for generations.
There is a moral difficulty here for affluent Western advocates of rapid global decarbonisation. It is easy for somebody sitting in an air-conditioned Australian office, surrounded by computers, refrigeration, electric lighting, telecommunications and modern transport, to tell developing populations that humanity must reduce its energy footprint. Indians and Chinese may reasonably respond that they intend to enjoy more of the prosperity that abundant energy made possible in the West.
Nor is the additional demand confined to developing countries. Just as politicians have committed advanced economies to electrifying transport, heating and industry, another enormous consumer has arrived: artificial intelligence.
AI does not live in some ethereal cyberspace. Every query, model-training run and inference operation ultimately requires physical machines housed in physical buildings consuming physical electricity. Those machines must be manufactured, powered and cooled.
The IEA estimates that global data-centre electricity consumption could more than double to approximately 945 TWh by 2030. Consumption by accelerated servers, driven principally by AI adoption, is projected to grow at around 30 per cent annually. In the United States, data centres are expected to account for roughly half the increase in electricity demand through 2030.
The irony is extraordinary. At precisely the historical moment when governments have committed themselves to electrifying transport, industry and heating in order to eliminate fossil fuels, the digital revolution has produced a vast new category of electricity consumption.
This does not mean renewable energy is useless. Renewable generation is expanding rapidly and will satisfy a substantial proportion of new electricity demand. China itself is building enormous quantities of solar and wind capacity. But China is also a useful demonstration of the difference between adding renewable energy and eliminating fossil fuels. When electricity demand grows rapidly enough, new renewable capacity can supplement existing energy sources rather than completely replace them.
The IEA's own analysis of Chinese data centres illustrates the problem. Coal presently supplies close to 70 per cent of their electricity. Between 2024 and 2030, coal generation supplying Chinese data centres is expected to increase by nearly 90 TWh annually, even while renewable generation serving them expands by approximately the same amount.
This is the arithmetic that political slogans cannot abolish. Building a gigawatt of renewable capacity is not equivalent to removing a gigawatt of fossil-fuel capacity if electricity demand is simultaneously increasing. Some new generation must first satisfy the additional load before it can displace anything that already exists.
Net zero consequently requires the world to win two races simultaneously. The first is to construct enough low-carbon generation, storage, transmission and associated infrastructure to satisfy rapidly increasing demand. The second is to construct still more capacity to replace the immense fossil-fuel system upon which the existing world economy depends.
And electricity is only part of the problem. Oil dominates transportation, natural gas provides industrial heat and feedstocks, coal remains deeply embedded in steelmaking and electricity production, while aviation, shipping, cement, fertilisers and heavy industry present additional decarbonisation measures. Electrifying some of these activities simply transfers their energy demand onto an electricity system that is already expanding.
None of this establishes that net zero violates the laws of physics. Given sufficiently cheap nuclear power, revolutionary energy storage, enormous transmission investment, carbon capture, technological breakthroughs and several decades of sustained political commitment, much deeper decarbonisation is technologically conceivable. But "technologically conceivable" and "actually going to happen" are very different propositions.
Net zero is ultimately a global claim. Australia could theoretically reduce its domestic emissions to almost nothing and have little influence upon atmospheric carbon dioxide if fossil-fuel consumption elsewhere continued growing. Climate physics does not recognise Australian borders. A tonne of carbon dioxide emitted in China or India enters the same atmosphere as one emitted in Melbourne.
This creates the fundamental political problem. Developing countries place economic development, industrialisation and rising living standards very high on their list of priorities. China also regards energy security and industrial capacity as strategic necessities. India has hundreds of millions of citizens whose per-capita energy consumption remains far below Western levels. Neither country is likely to sacrifice economic development merely so that Western governments can meet politically chosen dates on climate-policy calendars. And even if the hypothetical climate catastrophe was certain, probably neither country would change economic course and energy consumption in the slightest. Hence all of Australia's zero net fanaticism, is for nothing.
Meanwhile, Western countries themselves are discovering that decarbonisation becomes progressively harder as the easiest reductions are exhausted. Closing an old coal station and replacing some of its annual output with renewable generation is one thing. Maintaining a reliable electricity system dominated by intermittent generation while simultaneously electrifying cars, heating and industrial processes and connecting enormous AI data centres is quite another.
The danger is that net zero becomes less an engineering program than an exercise in political accounting. Domestic emissions can sometimes be reduced by shifting energy-intensive manufacturing overseas, after which finished products are imported from countries burning fossil fuels to manufacture them. The national carbon ledger improves while the atmosphere notices very little difference.
There is therefore a strong case for separating two questions that politicians habitually merge. The first is whether reducing greenhouse-gas emissions is desirable. The second is whether global net zero within currently nominated timelines is realistically achievable.
One can answer yes to the first without pretending that the second automatically follows. The numbers increasingly demand this distinction. China is adding electricity consumption on the scale of the European Union. India is industrialising rapidly. Developing Asia wants air conditioning, cars, appliances, factories and modern infrastructure. Western economies are attempting mass electrification. Artificial intelligence is creating gigantic new electricity loads. Oil, gas and coal consumption remain extraordinarily large.
The world is therefore attempting something unprecedented: replacing its dominant energy system while simultaneously demanding vastly more energy from it. Perhaps technological innovation will eventually solve the contradiction. Advanced nuclear reactors, cheaper storage, improved grids, geothermal power or technologies not yet commercially mature could radically change the equation. History warns against declaring technological problems permanently insoluble.
But policy should be based upon the world that exists rather than the world politicians hope will arrive. And the world that exists in 2026 is not rapidly abandoning energy consumption. It is demanding more of it. China wants more. India wants more. The developing world wants more. AI wants considerably more. Electrification itself demands more electricity.
Against that background, the relevant question is no longer simply whether net zero would be desirable. It is whether the global energy system is actually moving toward it. At present, the evidence suggests that the destination may be receding faster than governments care to admit. Zero net then is simply not going to happen, and from our position, it should never have been considered in the first place.
https://www.macrobusiness.com.au/2026/08/is-net-zero-even-possible/