At an inflection point where the world needs to reverse trends in emissions and energy consumption, AI is making it more difficult to meet climate commitments.
Over the weekend at COP29 in Baku, leaders endorsed a declaration to use digital tech and AI to accelerate climate action. However, key to this is curbing the emissions that result from the technology sector, reducing its manufacturing footprint, and tackling the growing problems of AI’s power appetite and e-waste.
At the first UN Climate Change Conference in 1995, digital technology was not on the agenda. Back then, the internet was used by less than 1% of the world’s population and the tech impact was negligible. By 2022, 166 tech companies generated nearly 1% of global greenhouse gas emissions, equivalent to one of the top 25 emitting countries. Ten of those companies accounted for almost 1% of global electricity consumption.
A single query via ChatGPT reportedly takes ten times the energy as a Google search.
Tech and AI are changing the climate equation. Greenhouse gas emissions from some of the largest tech companies have already risen by about 50% in five years. Their electricity consumption has more than doubled. Some analysis suggests emissions from Google, Microsoft, Meta and Apple may be 7.62 times higher than official tallies.
The energy generation and access requirements for AI are exponential. A single query via ChatGPT reportedly takes ten times the energy of a Google search – 2.9Wh vs 0.3Wh. If not addressed, the skyrocketing growth of AI and cloud computing will derail climate commitments as tech company net-zero roadmaps are shredded and data centres ratchet up energy supply requirements.
Google, for example, has dropped its long standing carbon-neutral promise, following a spike in emissions driven by its pursuit of AI, according to a 2024 environmental report. Since 2023, Google has no longer “maintained operational carbon neutrality” according to the report.
Keeping global warming within 1.5°C above pre-industrial levels – the tipping point for catastrophic climate change – requires deep, rapid reductions by countries, industries, communities, and individual consumers. Global emissions need to be cut by almost half by 2030, or 9% every year, yet they are still growing.
Despite pledges at COP28 a year ago, the burning of coal, oil and gas continued to rise in 2024 as emissions hit new highs.
UN Secretary-General António Guterres’ remarks at the opening of COP29 remind us of some progress. Last year – and for the first time – the amount invested in greens and renewables overtook the amount spent on fossil fuels. And almost everywhere, solar and wind are the cheapest sources of electricity. So, doubling down on fossil fuels is absurd.
The clean energy revolution is here, yet its transition is a geopolitical scramble to secure transformative technologies. Across the globe, renewables provide an increasing proportion of energy. However, they are not without their own issues and supply variability. Grid stability, baseload requirements and storage are part of the reason the controversial nuclear debate is raging in Australia ahead of the election. Microsoft, Google and Amazon are themselves looking to nuclear power to provide low emissions electricity as AI booms.
Tech and AI offer some opportunities in climate monitoring, mitigation and adaption. Climate monitoring has improved significantly using AI analysis of satellite imagery, for example in measuring iceberg change, deforestation, water management and wildlife conservation. Climate mitigation measures could include use of AI to optimise energy usage and even transform energy from commodity to technology. While climate adaption could benefit from AI-enabled early warning indicators.
So where to from here?
We need to reimagine our digital future to ensure it aids our climate efforts, rather than hinders them.
Big picture, the worst case is a retreat from multilateralism and a focus on sovereign security at the expense of climate commitments.
What this looks like is states focusing on strategic advantage for their own nation ignoring the current and future impacts. Examples might include elevating energy generation to “national security” status, along with water, needed in increasing volume to cool computing systems. Or the tech competition could further the exploitation of natural resources, clean water and arable land at the expense of biodiversity, ignoring climate goals.
President-elect Donald Trump is expected to withdraw from the Paris Agreement, again, and promote fossil fuels (“drill, baby, drill”). However, Trump, as always, is a wildcard – not all his rhetoric makes it to policy.
Undoubtedly, the United States is a key player (both as an emitter and fighter of climate change), however hundreds of nations have adopted the legally binding 2015 Paris Agreement. But the US government will probably sit this one out at a crucial inflection point for climate. Global leadership will have to come from somewhere else – others must guide the way.
The best-case scenario is radical change to stop catastrophic climate change, now. We are on the cusp of five catastrophic climate tipping points.
There is global public consensus. The largest ever public opinion survey on climate change, the Peoples’ Climate Vote 2024, shows 80% of people want their governments to take stronger action to tackle the climate crisis. More than 73,000 people, speaking 87 different languages across 77 countries were surveyed.
It has been argued that as Meta allows military agencies to access its AI software, it poses a moral dilemma for everybody who uses it. We may well soon be asking the same question in relation to climate. We need to reimagine our digital future to ensure it aids our climate efforts, rather than hinders them.
This post was originally published on here
