Insights into AI's influence on climate change won't prevent inertia in addressing the issue
In the quest to address the climate crisis, it's crucial to adopt a holistic approach that considers the impact of digital technology on the environment. This is a call to action, led by economist Mariana Mazzucato, who questions whether asking for more data evidence is truly addressing the climate crisis or just serving the interests of big tech.
One area of concern is the lack of transparency over energy consumption by tech companies. Google, for instance, has seen a 48% increase in its greenhouse gas emissions since 2019, and the tech giant has attributed its loss of confidence in meeting net-zero targets to the uncertainty around the future environmental impact of AI.
Mazzucato argues that big tech should not be let off the hook easily regarding their environmental impact. Beyond demanding data transparency, arguments for holding big tech accountable emphasise the need for standardised energy and environmental metrics, enforceable sustainability regulations, supply chain accountability, and integration of AI in resource governance.
Current self-reporting by tech firms is often voluntary, inconsistent, or misleading due to narrow metrics like PUE or reliance on carbon offsets that obscure true emissions. For example, AI’s real carbon footprint can be vastly underestimated (up to 662% higher) without standardised metrics.
Proposed solutions include congressional and federal executive action to direct agencies such as the Department of Energy (DOE), National Institute of Standards and Technology (NIST), Environmental Protection Agency (EPA), and others to develop and enforce comprehensive, uniform metrics covering all AI lifecycle stages (model training, inference, data center operation).
Regulatory norms beyond transparency should mandate minimum energy-efficiency standards, reliable renewable energy sourcing, and accountability for e-waste management and hardware lifecycle, since many data center operators fail to track or responsibly recycle retired equipment.
Supply chain transparency is critical. AI systems rely on minerals whose extraction often involves environmental degradation, corruption, inequality, and conflict. Accountability frameworks must incorporate ethical sourcing standards and independent international reporting requirements for big tech firms—enabling policymakers to intervene with evidence-based sustainable practices.
AI technology itself can aid environmental governance by monitoring illicit mining and resource conflicts via large-scale data, satellite imagery, and machine learning, assisting in enforcing sustainability standards and protecting vulnerable communities.
Industry initiatives like OpenAI and Google adopting renewable and carbon-neutral energy are positive but insufficient on their own without mandated rules and metrics, to avoid greenwashing and ensure real impact. Agreeing on standards, criteria, models, and methodologies for calculating and evaluating energy consumption and carbon emissions is necessary for sharing and accessing such data.
Digital carbon calculators, such as Green Software Foundation tools, Impact Framework, Green Algorithms, and Cloud Carbon Footprint, have proliferated, each approaching the measurement of carbon emissions differently. A focus on adaptable frameworks and bringing together different research communities is more effective in addressing the climate crisis than solely collecting and making data accessible.
In summary, the main arguments stress that transparency alone is insufficient, and a combination of standardised metrics, enforceable regulations, supply chain accountability, and using AI for resource governance are essential steps to hold big tech responsibly accountable for their AI environmental footprint beyond just making energy data public.
Science and technology are essential tools in the fight against climate change, but it's crucial to ensure that digital technology, particularly big tech, does not exacerbate the environmental crisis. The urgent need for standardized energy and environmental metrics, enforceable sustainability regulations, and supply chain accountability in the tech industry is emphasized, as current self-reporting is often inadequate or misleading. Furthermore, the integration of AI in resource governance could aid environmental protection, provided that ethical sourcing standards are incorporated and international reporting requirements are enforced.
