Who Decides the Environmental Cost of AI? A Case for Community Stewardship

Who Decides the Environmental Cost of AI? A Case for Community Stewardship

Artificial intelligence is often discussed as a digital technology, but its footprint is still profoundly physical. Behind every AI model are data centres consuming electricity, freshwater cooling systems, expanding energy grids, mining operations extracting critical minerals, and vast networks of infrastructure that occupy land and reshape local environments. And as AI continues to scale, these physical systems will increasingly influence not only how we work, but how we manage our natural resources.

Yet many of these decisions are still made through top-down processes by governments, technology companies, and infrastructure providers, with communities often consulted only after major decisions have already been made.

So, if AI is becoming part of our collective infrastructure, then decisions about its environmental footprint should not belong solely to tech companies with commercial interests. They should also involve the communities who will live with its impacts.

Below is one approach and framework for building a more participatory future of technology.

1. Prioritise Public Understanding

Communities cannot meaningfully participate in decisions they do not understand. While much of the public discussion around AI focuses on productivity, automation, and innovation, far less attention is given to the environmental systems that make these technologies possible. And if the environmental impact is raised, it’s often shown as the worst-case scenario, and doesn’t include the many regenerative solutions being developed.

People deserve accessible information about questions such as:

• How much energy do AI systems consume?

• Where are new data centres being built?

• How much freshwater is required for cooling?

• What minerals are needed to manufacture the underlying hardware?

• What are the trade-offs between technological growth and environmental sustainability?

• And what regenerative solutions are being developed for these issues?

Understanding these questions transforms AI from an abstract digital conversation into one about shared resources. And the more informed the public is, the more agency we have to build an equitable future.

2. Make Environmental Infrastructure Visible

One of the greatest challenges of digital technology is that its physical infrastructure is largely invisible. Most people never see the data centres powering their searches, the transmission lines supporting computational demand, or the mining operations supplying semiconductor manufacturing… all until they’re being built in their backyard without consultation.

Making this infrastructure visible (before building them) through public reporting, environmental impact assessments, interactive maps, and open-access education enables communities to better understand the relationship between technological progress and our ecological systems.

Invisible infrastructure often escapes public conversation, but visible infrastructure invites stewardship.

3. Build Participatory Governance

Environmental decisions have long benefited from public consultation. As AI infrastructure expands, that principle should extend to technological development. Communities could participate through citizen assemblies, local advisory panels, digital consultation platforms, and public deliberation processes when significant technology infrastructure is proposed.

Questions might include:

• Should new data centres be approved in water-stressed regions?

• What environmental standards should technology companies meet?

• How should renewable energy be prioritised?

• What environmental offsets should accompany major infrastructure projects?

The goal is not to slow innovation, but to ensure innovation develops with social legitimacy and environmental responsibility.

4. Define Shared Environmental Boundaries

Not every technological advancement carries the same environmental cost, and communities should help determine where environmental boundaries should exist before infrastructure becomes embedded.

This might include agreements around:

• Freshwater use for cooling systems.

• Renewable energy requirements for high-compute facilities.

• Biodiversity protections around new infrastructure.

• Responsible sourcing of critical minerals.

• Circular economy commitments for electronic waste.

Boundaries established early are far easier to uphold than those introduced after systems have already scaled.

5. Recognise That Technology Is Ecological & Diverse

One of the greatest misconceptions of the digital age is that technology exists separately from nature. But in reality, every algorithm relies on physical materials, energy systems, water, land, and human labour. And it’s important to remember that AI and emerging digital technologies are not replacing the natural world, but becoming another layer within it.

Recognising this changes the conversation from technological optimisation to ecological stewardship. So, conversations about AI infrastructure cannot be led exclusively by engineers and technology companies. We need diversity in discussions to understand the broader impact.

From ecologists who understand ecosystems; Anthropologists who understand communities and cultural adaptation; Indigenous knowledge holders who understand long-term relationships with place; Urban planners who understand infrastructure; Economists who understand incentives; Environmental scientists who understand planetary boundaries. And our communities themselves who understand what is worth protecting.

The future of AI will be strongest when these perspectives are considered together rather than in isolation.

A Future That Balances Innovation With Stewardship

The environmental footprint of AI is still being shaped. That means we have an opportunity to influence its trajectory before today's infrastructure becomes tomorrow's default. If we want technological progress that strengthens both society and the environments we depend upon, then communities must become active participants in shaping how these systems are built.

That begins by:

• Expanding public understanding of AI's environmental footprint.

• Making digital infrastructure more transparent.

• Creating meaningful opportunities for community participation.

• Establishing environmental boundaries before systems scale.

• Bringing together expertise from technology, environmental science, the humanities, and local communities.

The question is no longer whether AI will reshape our future. It is whether we will collectively help shape the environmental future that AI creates.

We dive deeper into the relationship between technology and nature in this podcast episode: