By: Norman Valdez, Alex Tveit and Manuela Garzon

Picture a powerful technology that could help us predict weather patterns with unprecedented accuracy and optimize our cities’ energy use down to the last kilowatt. That’s the promise of artificial intelligence (AI). But like any groundbreaking innovation, AI comes with its own environmental story – one that’s more complex than many realize.

AI comes with both opportunities and challenges for environmental sustainability. On one side, it’s helping scientists develop smarter climate models and engineers create more efficient power grids. On the other, training these sophisticated AI systems – especially the chatbots and language models we’re increasingly relying on – demands enormous amounts of electricity and water, leaving a substantial carbon footprint that rivals that of some small countries.

Here’s the added twist that often goes unnoticed: while AI’s environmental costs affect us all, they hit hardest in communities already struggling with the effects of the climate crisis. As an example, a data center in a drought-prone region, consuming millions of gallons of water while local farmers watch their crops wither.

As we race to embrace AI’s possibilities, we face a crucial challenge: how do we harness its potential to solve environmental problems while ensuring we’re not making those very problems worse?

The Opportunity

AI can be a powerful ally in mitigating environmental risks and enhancing sustainability. Look at Alpha El, an Indigenous-led AI company that harnesses thermal imaging, vision, and satellite mapping together with AI models to detect wildfires in their early stages. This initiative can help prevent the spread of fires that continue to affect communities from coast to coast to coast. Wildfires not only destroy vast tracts of forest and ecosystems but also impose heavy economic burdens. In a single week in June 2023, heavy smoke from wildfires cost Ontario over $1.2 billion in health impacts, including premature deaths and increased hospital visits.

In sustainable agriculture, AI-driven precision farming techniques enable farmers to monitor soil health, optimize water usage, and reduce pesticide application, thereby decreasing environmental strain. AI-powered drones and sensors can detect crop diseases early, ensuring timely interventions that save resources and improve yields. Additionally, AI aids in biodiversity conservation through species monitoring and habitat mapping, and through organizations like Crowdsorsa, the public and AI models are working together to help communities remove invasive species. 

By leveraging AI’s capabilities in these areas, we can not only mitigate the immediate impacts of climate change but also promote long-term environmental resilience. These applications illustrate how AI, when thoughtfully deployed, serves as a catalyst for positive environmental transformation, demonstrating that technology can be aligned with the principles of sustainability and equity.

The Challenges

While AI holds immense promise, its environmental toll is undeniable. Training large AI models, such as GPT-4, demands an enormous amount of energy, contributing significantly to carbon emissions. For instance, training a single state-of-the-art AI model can emit as much carbon dioxide as five cars in their lifetimes. As AI technology advances, these energy demands—and their carbon footprints—are projected to grow exponentially, with the ICT industry expected to account for 14% of global emissions by 2040.Data centers powering AI systems rely on substantial volumes of water for cooling, placing strain on freshwater resources. Google’s data centers used 20% more water in 2022 than in 2021, while Microsoft’s water use rose by 34%. This increases resource competition and can exacerbate local water shortages, particularly in drought-prone areas. To address these challenges, the industry is exploring solutions such as geographical load balancing, renewable energy adoption, and improving AI model efficiency. These efforts reflect a growing recognition within the tech sector of the need to balance innovation with environmental stewardship.

Policy and Advocacy

The path toward environmentally responsible AI requires concerted action from multiple stakeholders. Looking at Canada as a case study reveals both progress and challenges in developing comprehensive AI environmental policies.

Canada’s approach to AI policy and advocacy illustrates how nations are grappling with balancing innovation and environmental responsibility. The federal government has taken important steps toward engaging stakeholders in discussions about responsible AI development. In 2019, they established the Advisory Council on Artificial Intelligence, tasked with identifying economic growth opportunities while ensuring ethical AI development. Following this, a Public Awareness Working Group was created in 2020 to gather national perspectives on AI’s potential and challenges. Through national surveys and virtual workshops, the government has worked to understand public sentiments about AI across the country.

However, the gap between ambition and action remains significant. Despite allocating $2.4 billion to support the AI sector in the 2024 federal budget, the initiative notably lacks specific measures to address AI’s growing environmental footprint. Environmental advocates are pushing for stronger policies to tackle the substantial energy consumption, water usage, and e-waste associated with AI development. 

Shifting governance frameworks to integrate various perspectives and address the environmental impact of artificial intelligence is essential. For example, Indigenous knowledge systems have the potential to align with provincial and federal development of policies, to emphasize a holistic approach to environmental governance, which will support generations to come. 

Collaboration between the government and citizens is essential, as when collaboration takes place diverse perspectives are reflected while integrating knowledge systems. It is absolutely critical to develop ethical statements, standards, education, training and digital infrastructure in Canadian AI governance. Approaching future challenges through a collaborative lens is essential for protecting ecosystems and, by working alongside Indigenous perspectives, the government can build ethically grounded AI environmental frameworks. As Canada strives to establish itself as a leader in AI innovation, the challenge of balancing technological advancement with environmental stewardship becomes increasingly pressing.

Looking Forward

The relationship between AI and environmental sustainability represents one of the most pressing challenges of our time. As Canada’s experience demonstrates, even nations at the forefront of AI development struggle to strike the right balance between innovation and environmental responsibility. While AI offers unprecedented opportunities to address environmental challenges, we must ensure these benefits don’t come at too high an ecological cost.

The gap between technological ambition and environmental action needs urgent attention. Success will require meaningful collaboration between technologists, environmentalists, policymakers, and communities to integrate environmental considerations into every aspect of AI development. By implementing robust policies and maintaining accountability, we can harness AI’s potential while protecting our planet’s resources. The choices we make today will determine whether AI becomes a tool for environmental regeneration or another contributor to our ecological challenges.

Cover image generated with Midjourney.