The Environmental Impact of Artificial Intelligence Data Centers
- Amazon Connection Carbon
- May 24
- 2 min read
Introduction
Artificial Intelligence (AI) has transformed entire sectors of the economy and science, but its rapid growth comes with an often-overlooked environmental cost. The infrastructure required to train and operate sophisticated algorithms consumes increasing amounts of electricity and natural resources. The data centers that support these operations have become true invisible carbon emitters.

Energy Consumption in Numbers
According to a study by the University of Massachusetts, training a single large AI model can emit over 284 tons of CO₂—the equivalent of about five times the carbon footprint of an average car over its entire lifetime.
Currently, data centers are responsible for approximately 2% to 3% of global electricity consumption. With the rapid expansion of generative AI and real-time usage by billions of users, this figure could double by 2030—making the sector one of the top emitters within the digital industry.
Beyond energy use, data center infrastructure requires massive volumes of water for cooling, especially in hotter regions. This demand can directly affect the water availability of nearby urban and rural areas.
Why This Should Concern Brazil (and the Amazon)
Brazil is among the countries with the greatest potential to host data centers due to its vast territory, access to renewable energy, and growing digitalization. However, this advantage will only remain sustainable if environmental impacts are effectively managed.
In the Amazon, unregulated expansion of large-scale infrastructure could increase pressures on land, water, and biodiversity. Without effective mitigation and compensation measures, indirect impacts could undermine climate commitments and conservation targets.
How ACC Views This Challenge
At Amazon Connection Carbon (ACC), we believe innovation must go hand in hand with environmental responsibility. We encourage technology and infrastructure companies to assess their carbon footprint and adopt compensatory measures that make a real difference—such as purchasing carbon credits from high-integrity projects like REDD+.
ACC's projects in the Amazon contribute to balancing emissions and maintaining essential ecosystem services, such as water regulation and carbon storage. We also offer customized solutions for corporate emission offsets that include added social impact.
Conclusion
Artificial intelligence can be a powerful ally in the fight against climate change—provided its own environmental impacts are acknowledged and responsibly offset.
Enjoyed the content?
Stay aware of the invisible impacts of technology and follow ACC’s work for innovation aligned with the preservation of the Amazon.
References
Exame – “IA não pode ser sustentável se o impacto dos data centers não for reconhecido”: https://exame.com/bussola/ia-nao-pode-ser-sustentavel-se-o-impacto-dos-data-centers-nao-for-reconhecido/
UMass Amherst – Energy and Policy Considerations for Deep Learning in NLP: https://arxiv.org/abs/1906.02243
International Energy Agency (IEA) – Data Centres and Data Transmission Networks: https://www.iea.org/reports/data-centres-and-data-transmission-networks
Comments