More than 200 laws regulating artificial intelligence have been enacted across over 100 countries, yet most overlook a critical problem: the massive environmental toll of AI systems. While regulators focus on privacy, bias, and security, they're largely ignoring the greenhouse gas emissions, water consumption, and e-waste generated by AI development and deployment. A new analysis of regulatory frameworks in the European Union and United Kingdom reveals a troubling gap that could undermine climate goals as AI adoption accelerates .

What Environmental Damage Is AI Actually Causing?

The environmental impact of AI spans the entire lifecycle of the technology, from the moment rare earth elements are extracted for computer chips to the final disposal of hardware. Training large language models like GPT-3 consumes staggering amounts of resources. Researchers estimated in 2025 that training GPT-3 alone required approximately 700,000 liters of freshwater for electricity generation and data center cooling . The manufacturing of graphics processing units (GPUs), which are specialized computer chips essential for AI training, relies on extracting rare earth elements that can contaminate soil and water, pollute air, and destroy forest habitats.

The problem intensifies as AI models grow larger and more widespread. Even though individual AI models are becoming more energy efficient through advances in algorithms and optimization, the overall energy consumption continues rising because AI is being deployed across more industries and use cases. Critically, the energy consumed when people actually use AI systems, such as generating text or images, vastly outweighs the energy used during the initial training phase .

Why Are Current AI Laws Failing to Address Environmental Concerns?

The European Union's AI Act, which became the world's first comprehensive AI regulation when it took effect on August 1, 2024, acknowledges some environmental consequences and requires that AI systems be developed and used in a sustainable manner. However, the requirements lack teeth. The law asks AI providers to disclose energy consumption data, but only when specifically requested by the EU's AI Office. Additionally, codes of conduct to minimize environmental impact are optional, not mandatory .

The situation is even more dire in the United Kingdom, which has no AI-specific legislation. The UK government's 2023 white paper on AI regulation explicitly excludes sustainability from its scope, stating that environmental issues are "outside of the scope of our proposals for a new overarching framework for AI regulation," even while acknowledging that sustainability is an important consideration .

This regulatory blindness stems partly from a fundamental design choice: most AI laws are intentionally anthropocentric, meaning they prioritize human interests and wellbeing over environmental protection. The EU's AI Act, for instance, explicitly states that AI should serve as a tool for people with the aim of increasing human wellbeing, leaving environmental safeguards secondary.

How Can Regulators Make AI More Sustainable?

Experts propose several concrete measures that governments could implement to address AI's environmental footprint:

• Mandatory Transparency Requirements: AI developers should be required to disclose energy consumption, water usage, carbon emissions, rare earth elements extracted, and plastic used during production. This baseline data would enable policymakers to set meaningful targets and limits.
• Energy Efficiency Labeling Systems: An "energy star" rating system similar to the EU's existing appliance efficiency labels could rank AI systems from most to least energy efficient, helping consumers make informed choices about which AI tools to use.
• Consumer Warnings: Labels could display the carbon dioxide emissions or water consumed per query, making the environmental cost of AI use visible to end users.
• Financial Incentives: Tax breaks and funding incentives could encourage technology companies to invest in more sustainable AI development and deployment practices.
• Certification Programs: Social and environmental certification systems could provide third-party verification of AI sustainability claims.

These measures would give companies concrete targets to work toward while providing consumers with the information needed to demand greener AI alternatives .

Are Concerns About AI Energy Demand Overblown?

While the environmental challenges are real, some experts argue that popular narratives about AI consuming unlimited energy may be misleading. Safak Yucel, an associate professor of operations and associate director at Georgetown University's Business of Sustainability Initiative, explains that the relationship between AI growth and energy demand is more nuanced than headlines suggest .

"Yes, there will be more AI use, but AI is also becoming much more efficient in terms of electricity and water use. Greater AI use may not translate into significantly higher electricity demand," explained Safak Yucel, associate professor of operations at Georgetown University's McDonough School of Business.

Safak Yucel, Associate Professor of Operations, Georgetown University

Advances in basic sciences like statistics and optimization are enabling AI algorithms to deliver better results with less computational power. Google reported that emissions for its Gemini model dropped by a factor of 44 over one year, demonstrating that efficiency gains are possible at scale .

Moreover, increased electricity demand from AI doesn't necessarily require fossil fuels. Renewable energy sources like solar and wind can meet growing power needs if investments continue. Over the past decade, coal's share of the US electricity grid fell from 40% to about 14%, while renewables jumped from 10% to about 23%, largely because renewables have become cheaper than coal .

Battery storage technology offers another solution. By charging batteries during peak solar generation hours and using that stored electricity when the sun isn't shining, grids can balance renewable energy's intermittency without relying on conventional power plants .

What Should Happen Next?

The gap between AI's rapid expansion and regulatory oversight is widening. Without mandatory environmental standards built into AI laws, companies have little incentive to prioritize sustainability over performance or cost. The EU's AI Act represents progress, but its voluntary elements and limited enforcement mechanisms fall short of what's needed.

Policymakers in the UK, US, and other countries should learn from these shortcomings and integrate environmental accountability directly into AI regulation from the start. By requiring transparency, setting efficiency targets, and giving consumers visibility into AI's environmental costs, regulators can align AI development with climate goals rather than working against them. The technology is advancing rapidly; the regulatory framework must keep pace .