Unveiling the Environmental Costs of AI: The Role of Tax Policy in combating Climate Change

1. 1.     The rise of AI emissions 

In recent years, the environmental toll of AI has surged to alarming levels. In 2024, Microsoft reported total market-based emissions (encompassing Scope 1, 2, and 3 emissions) of 17.16 million metric tons of CO2,[1]while Google’s greenhouse gas emissions stood at 14.31 million metric tons of CO2.[2] However, a recent investigation by The Guardian[3] suggests that the actual environmental impact of AI may be far greater than these numbers imply. Their analysis reveals that emissions from the in-house data centers of companies like Google and Microsoft could be up to 662% (or 7.62 times) higher than officially reported.[4] This stark difference highlights the underreporting of energy consumption and emissions tied to AI operations.

One of the central challenges in accurately assessing and reporting environmental costs lies in the use of creative accounting practices and regulatory loopholes that obscure true emissions figures. Creative accounting refers to manipulating reports to present a more favorable image while remaining within legal bounds. In the tech industry, a prime example is the use of Renewable Energy Certificates (RECs).[5]

By purchasing RECs, companies can claim they are using renewable energy, even if the electricity powering their facilities comes from non-renewable sources. These certificates allow tech giants to report market-based emissions, which look more favorable than the actual location-based emissions – those generated by the local energy grids their data centers rely on.[6]

When RECs are excluded, the real emissions from data centers can be significantly higher. In fact, if tech companies’ data centers were ranked solely by their location-based emissions, they would rank as the 33rd highest emitters in the world, competing with countries such as the Philippines and Algeria.[7]

This practice illustrates how creative accounting can mislead stakeholders about the true environmental impact of large corporations.

The controversy surrounding RECs has sparked debate among key players in AI development. Google, for instance, has announced plans to move away from reliance on RECs, proposing a “24/7” model that would ensure energy consumption is matched with clean energy from the same grid at the same time.[8] Microsoft, while supporting a shift toward more location- and time-specific energy use, has yet to fully back Google’s approach.[9]

On the other hand, Amazon and Meta, along with the Emissions First Partnership,[10] advocate for greater flexibility in REC use, allowing companies to purchase certificates regardless of geographic limitations. They argue that this flexibility is crucial for driving real-world emission reductions and encouraging renewable energy investment, especially in regions with restrictive foreign investment policies.[11]

1. 2.     The European Approach to climate-friendly AI technologies

As economic and political actors debate the merits and shortcomings of RECs in climate reporting, the environmental costs of AI continue to rise, exacerbating the climate crisis. This calls for immediate policy intervention to address the growing impact of new technologies on the environment.

In this context, the EU Commission is exploring measures to improve energy efficiency and circular economic performance in cloud computing and data centers to reach a compromise in the race between climate change and the digital revolution.[12]

These efforts follow a study on Energy-efficient Cloud Computing Technologies and Policies for an Eco-friendly Cloud Market, which addresses the issue of growing energy consumption due to the expansion of cloud services in Europe.[13]

This study proposes both technological and policy-based solutions to reach efficient levels of energy consumption in cloud computing. A key focus of the study is the potential of Green Public Procurement (GPP)[14] in promoting energy-efficient cloud services. In this sense, the study recommended the implementation of GPP criteria into national action plans. 

Another key focus of the study is the importance of research and technological development (RTD) for improving cloud computing efficiency. In total, the study formulates 38 different RTD recommendations across many sectors aimed at reducing energy consumption and greenhouse gas emissions.[15] These 38 recommendations were narrowed down to seven key RTD policy recommendations through stakeholder workshops. With these recommendations, the European Commission is considering implementing them in a ‘cloud rulebook’ to establish a coherent framework for cloud services.[16]

Following this study, the Commission set a plan aimed at reaching its environmental goals in the field of cloud computing and AI. In doing so, the Commission will rely on a mix of existing instruments, reviews of existing legislation and new regulatory initiatives. 

Existing measures include the EU Code of Conduct on Data Centre Energy Efficiency,[17] which is a voluntary agreement that encourages data center operators to improve energy efficiency by implementing best practices and benchmarking their performance. Moreover, the Ecodesign Directive 2009/125/EC[18] establishes a framework for setting minimum energy efficiency requirements for energy-related products. Lastly, the European Directive on Energy Efficiency 2012/27/EU[19] provides binding measures to help the EU achieve its energy efficiency targets. 

The study on cloud computing suggests undertaking reviews of these existing EU legislations to address the problem of energy consumption.[20] The Ecodesign Directive, which currently focuses on physical products like servers and data storage devices, could be reviewed to assess the feasibility of expanding its scope to include the service component of cloud computing.[21] Moreover, it also suggests a review of the existing EU legislation to establish minimum criteria for the energy efficiency of newly built data centers. This could be done by amending the Energy Efficiency Directive or the Commission Regulation on servers and data storage products to include provisions for minimum PUE values.[22]

1. 3.     The role of tax policy in leading sustainable AI development

Following the recent undertakings and efforts of the European Union, The International Monetary Fund (IMF)[23] recently published a position paper recommending a targeted tax on data centers as an effective way to reduce emissions.[24]

Data centers and crypto mining already account for 2% of global electricity usage, a figure expected to increase to 3.5% by 2027, this makes their electricity consumption equivalent to the current electricity consumption of Japan. To put this in an even broader perspective Japan is the fifth largest electricity consumer in the world.[25]

The IMF estimates that a tax of $0.032 per kilowatt hour on data centers, or $0.052 including air pollution costs, could generate $18 billion annually while incentivizing the adoption of energy-efficient technologies and cleaner energy sources. On the other hand, the IMF has proposed a different rate on crypto mining of $0.047 per kilowatt hour, or $0.089 including air pollution costs. The reason for the difference in price between the data centers and the crypto mining stems from the tendency of data centers to be located in areas with access to greener electricity sources compared to crypto mining operations.[26]

The current situation is highlighted in the IMF paper where data centers and crypto miners are benefiting from generous tax exemptions and incentives on income, consumption, and property. Considering the environmental damage caused by these industries, the limited employment opportunities they generate, the strain they place on electrical grids and the environmental impact of AI emissions, the net benefits of these special tax regimes are questionable. It is highlighted in this paper that the pressure that these industries place on the electrical grids could lead to higher prices for households and reduced demands for other low-emission goods, such as electric vehicles.[27]

Because of this the IMF underscores the urgency of adopting appropriate carbon pricing and that renewable energy sources be expanded to mitigate the effects of climate change. The targeted taxation of data centers and crypto mining, alongside measures like credits for zero-emission technologies and bilateral power purchase agreements, could help curb rising emissions.[28]

Considering the growing environmental impact of AI and the ineffectiveness of existing policies, the proposal of the IMF could mark a crucial step in mitigating the environmental impact of AI development and ensuring a more sustainable future. 

Still, proposals such as the position paper of the IMF may undermine the development of AI technologies, by placing an unreasonable cost on the developers and providers of AI systems. 

As the trade-off between AI development and climate change keeps growing, policymakers are urged to find solutions that can serve both the digital transition and the race towards a sustainable economy.

Domestic and international tax policy, in its various forms and structures, can play a key role in this regard. As exemplified in the example of the IMF, the adoption of Pigouvian taxes that constrain the development of AI could prove to be an effective – albeit controversial – method to reach a trade-off between digital progress and sustainability.