Artificial Intelligence and Sustainability: The Elephant in the Room
We look at the role of Artificial Intelligence and Sustainability, as AI revolutionises industries from healthcare to logistics.
AI is driving a generational shift in how we use technology. Its application is growing fast, and the questions we need to ask about what that growth means - for the environment, for communities, for the planet - are growing just as quickly.
The uncomfortable truth is that AI has the potential to both accelerate and undermine the sustainability goals we need to hit. It’s not one or the other. It’s both, simultaneously, and the balance depends entirely on the choices being made right now.
The energy question.
AI requires enormous computational power, particularly for training large models. That demands significant energy, often sourced from non-renewable grids. As adoption accelerates, so does the energy footprint.
There are positive signals though. On-device AI processing - pioneered by companies like Apple and Samsung - decentralises computation away from massive data centres. For tasks like voice recognition and image processing, this can meaningfully reduce the energy needed. But these efficiencies don’t cancel out the growth in demand. They just slow the curve.
The indirect impacts.
Beyond direct energy use, AI’s environmental influence is more complex. On one hand, AI enhances the efficiency of renewable energy systems - Google’s DeepMind work on wind power forecasting increased the value of wind energy by roughly 20%. Precision farming, powered by AI, can reduce chemical runoff and resource waste.
On the other hand, AI-driven mechanisation in agriculture can increase fossil fuel consumption if not managed responsibly. And AI’s ability to optimise manufacturing can just as easily increase output as reduce waste - it depends entirely on what it’s being optimised for.
Understanding the footprint.
We need better scenarios for how AI will affect different sectors. Conservative models that assume limited adoption. Ambitious models that align with the UN Sustainable Development Goals. And realistic models that account for how messy and uneven the transition will actually be.
Building these scenarios requires new data, new analytical frameworks, and cross-sector collaboration. Given the pace of change, they’ll also need constant updating. No single model will capture the full picture.
Will AI improve sustainability?
That’s the central question, and the honest answer is: it depends. The pace of change makes it genuinely hard to predict. What’s clear is that sustainability considerations need to be built into AI development and deployment from the start, not retrofitted afterwards.
At Human Kind, we monitor the latest research, publications, and forums to maintain an informed view of how AI impacts the environment at both micro and macro levels. We build those considerations into the work we advise on and the products we help create. Because the question isn’t whether to use AI. It’s whether we’re using it responsibly.
If you want to explore how AI can support your business goals without compromising on sustainability, that’s exactly the balance we help organisations find. Learn more about our Digital Product & AI work.
Further Reading
- Beyond the Hype: How Can AI Deliver Value to Industry
- What is Scope 1, 2 and 3?
- How to Build Efficient, Sustainable Digital Products