The environmental impacts of artificial intelligence (AI) may vary significantly. Many deep learning methods have significant carbon footprints and water usage. [1] Some scientists have suggested that artificial intelligence may provide solutions to environmental problems.
AI has a significant carbon footprint due to growing energy usage, especially due to training and usage. [2] [3] Researchers have argued that the carbon footprint of AI models during training should be considered when attempting to understand the impact of AI. [4] One study suggested that by 2027, energy costs for AI could increase to 85–134 Twh, nearly 0.5% of all current electricity usage. [5] [1] Training one deep learning model may use up to the same carbon footprint as the lifetime emissions of 5 cars. [2] Training large language models (LLMs) and other generative AI generally requires much more energy compared to running a single prediction on the trained model. [6] Using a trained model repeatedly, though, may easily multiply the energy costs of predictions. [6] The computation required to train the most advanced AI models doubles every 3.4 months on average, leading to exponential power usage and resulting carbon footprint. [7] Additionally, artificial intelligence algorithms running in places predominately using fossil fuels for energy will exert a much higher carbon footprint than places with cleaner energy sources. [8] These models may be modified for less environmental impacts at the cost of accuracy, emphasizing the importance of finding the balance between accuracy and environmental impact.
BERT, a generative AI model trained in 2019, consumed "the energy of a round-trip transcontinental flight." [9] GPT-3 released 552 metric tons of carbon dioxide into the atmosphere during training, "the equivalent of 123 gasoline-powered passenger vehicles driven for one year". [9] [10] [11] Much of the energy cost is due to inefficient model architectures and processors. [9] One model named BLOOM, from Hugging Face, trained with more efficient chips and, therefore, only released 25 metric tons of CO2. [10] Incorporating the energy cost of manufacturing the chips for the system doubled the carbon footprint, to "the equivalent of around 60 flights between London and New York." [10] Operating BLOOM daily was estimated to release the equivalent carbon footprint as driving 54 miles. [10]
Algorithms which have lower energy costs but run millions of times a day can also have significant carbon footprints. [10] The integration of AI into search engines could multiply energy costs significantly, [9] [12] with some estimates suggesting energy costs rising to nearly 30 billion kWh per year, an energy footprint larger than many countries. [13] Another estimate found that integrating ChatGPT into every Google search query would use 10 tWh each year, the equivalent yearly energy usage of 1.5 million European Union residents. [12]
Increased computational demands from AI caused both increased water and energy usage, leading to significantly more demands on the grid. [14] Due to increased energy demands from AI-related projects, coal-fired plants in Kansas City [15] and West Virginia [1] pushed back closing. Other coal-fired plants in the Salt Lake City region have pushed back retirement of their coal-fired plants by up to a decade. [16] Environmental debates have raged in both Virginia and France about whether a "moratorium" should be called for additional data centers. [15] In 2024 at the World Economic Forum, OpenAI executive Sam Altman gave a speech in which he said that the AI industry can only grow if there is a major technology breakthrough to increase energy development. [17] [18] [19]
In 2024, Google failed to reach key goals from their net zero plan as a result of their work with AI, [20] [21] and had a 48% increase in greenhouse gas emission attributable to their growth in AI. [14] [1] Microsoft and Meta had similar increases in their carbon footprint, similarly attributed to AI. [1] Carbon footprints of AI models depends on the energy source used, with data centers using renewable energy lowering their footprint. [7] Many tech companies claim to offset energy usage by buying energy from renewable sources, though some experts argue that utilities simply replace the claimed renewable energy with increased non-renewable sources for their other customers. [16] Analysis of the carbon footprint of AI models remains difficult to determine, as they are aggregated as part of datacenter carbon footprints, and some models may help reduce carbon footprints of other industries, [22] or due to differences in reporting from companies. [23]
Some applications of ML, such as for fossil fuel discovery and exploration, may worsen climate change. [4] [11] Use of AI for personalized marketing online may also lead to increased consumption of goods, which could also increase global emissions. [11]
AI chips, (i.e. GPUs) use more energy and emit more heat than traditional CPU chips. [1] AI models with inefficiently implemented architectures, or trained on less efficient chips may use more energy. [9] Since the 1940's the energy efficiency of computation has doubled every 1.6 years. [24] Some skeptics argue that improvements of AI efficiency may only increase AI usage and therefore carbon footprint due to Jevons paradox. [22]
In September 2024, Microsoft announced an agreement with Constellation Energy to re-open the Three Mile Island nuclear power plant to provide Microsoft with 100% of all electric power produced by the plant for 20 years. Reopening the plant, which suffered a partial nuclear meltdown of its Unit 2 reactor in 1979, will require Constellation to get through strict regulatory processes which will include extensive safety scrutiny from the US Nuclear Regulatory Commission. If approved (this will be the first ever US re-commissioning of a nuclear plant), over 835 megawatts of power – enough for 800,000 homes – of energy will be produced. The cost for re-opening and upgrading is estimated at $1.6 billion (US) and is dependent on tax breaks for nuclear power contained in the 2022 US Inflation Reduction Act. [25] The US government and the state of Michigan are investing almost $2 billion (US) to reopen the Palisades Nuclear Reactor on Lake Michigan. Closed since 2022, the plant is planned to be reopened in October 2025. The Three Mile Island facility will be renamed the Crane Clean Energy Center after Chris Crane, a nuclear proponent and former CEO of Exelon who was responsible for Exelon spinoff of Constellation. [26]
In 2025, Microsoft unveiled plans to invest $80 billion in the development and expansion of data centers designed to support AI technologies. These facilities, critical to the advancement of AI, depend on vast networks of interconnected chip clusters and significant electrical power to operate efficiently. [27]
In 2024, a US public policy group reported that AI and other technologies and industries poised to dominate the global economy are characterized by their high electricity demands. As such, the foundation of US energy strategy and policymaking will be to prioritize the reliable and abundant provision of electricity to support these critical sectors, which are needed to maintain the US economic and technological leadership in the twenty-first century. [28] The rapid proliferation of AI has created unprecedented demand for electrical power, presenting a major obstacle to the sector’s growth. E.g., in Northern Virginia, the largest global hub for AI data centers, the timeline for connecting bigger facilities—those requiring over 100 megawatts of power—to the electrical grid has extended to seven years, highlighting the strain on the energy infrastructure and the challenge of meeting AI’s escalating power needs. Across the United States, utilities are experiencing the most substantial surge in electrical demand in decades. This strain is directly contributing to longer wait times for grid connections, complicating efforts to maintain the country’s technological leadership in AI. The significance of these energy challenges extends beyond logistics. [29] A New York Times editorial emphasized the critical role of energy infrastructure, stating that “Electricity is more than just a utility; it’s the bedrock of the digital era. If the United States truly wants to secure its leadership in A.I., it must equally invest in the energy systems that power it.” [30]
Cooling AI servers can demand large amounts of fresh water which is evaporated in cooling towers. [22] [23] In fact, data centers housing AI are globally expected to consume six times more water than the country of Denmark. [31] By 2027, AI may use up to 6.6 billion cubic meters of water. [32] One professor has estimated that an average session on ChatGPT, with 10–50 responses, can use up to a half-liter of fresh water. [22] [33] [34] Training GPT-3 may have used 700,000 liters of water, equivalent to the water footprint of manufacturing 320 Tesla EVs. [33]
One data center that Microsoft had considered building near Phoenix, due to increasing AI usage, was likely to consume up to 56 million gallons of fresh water each year, equivalent to the water footprints of 670 families. [32] Microsoft may have increased water consumption by 34% due to AI, while Google increased its water usage by 20% due to AI. [34] [7] Due to their Iowa data center cluster, Microsoft was responsible for 6% of the freshwater use in a local town. [34]
E-waste due to production of AI hardware may also contribute to emissions. [7] The rapid growth of AI may also lead to faster deprecation of devices, resulting in hazardous e-waste. [35] Some applications of AI, such as for robot recycling, may reduce e-waste. [36] [37]
A deal was approved on September 20th, 2024 between Microsoft and Constellation energy to re-open the Three Mile Island nuclear plant. Bill Gates has stated Microsoft intends to use the power to power its usage of Open AI's services within its systems as well as peoples homes. [38]
Key regulatory permits for the plant's new life, however, haven't been filed, regulators say. [38] If the deal goes through, Three Mile Island would provide Microsoft with the energy equivalent it takes to power 800,000 homes, or 835 megawatts. This would be the first time a decommissioned plant would be reopened and the first time a commercial plant provided to a singular customer. [39]
AI has significant potential to help mitigate effects of climate change, such as through better weather predictions, disaster prevention and weather tracking. [40] [41] Some climate scientists have suggested that AI could be used to improve efficiencies of systems, such as renewable-energy systems. [13] Google has claimed AI could help mitigate some effects of climate change such as predicting floods or making traffic more efficient. [21] Some algorithms may help predict the impacts of more severe hurricanes, measure the melting of polar ice, deforestation, and help monitor emissions from sources. [11] [41] One machine learning project, the Open Catalyst project, has been used to identify "suitable low-cost electrocatalysts" for battery storage of renewable energy sources. [4] AI may also improve the efficiencies of supply chains and productions for environmentally detrimental industries such as food and fast fashion. [40] However, as yet there are no widely accepted frameworks which evaluate AI systems' total climate impacts, factoring in both costs and benefits. [42]
Electricity generation is the process of generating electric power from sources of primary energy. For utilities in the electric power industry, it is the stage prior to its delivery to end users or its storage, using for example, the pumped-storage method.
Electric power systems consist of generation plants of different energy sources, transmission networks, and distribution lines. Each of these components can have environmental impacts at multiple stages of their development and use including in their construction, during the generation of electricity, and in their decommissioning and disposal. These impacts can be split into operational impacts and construction impacts. All forms of electricity generation have some form of environmental impact, but coal-fired power is the dirtiest. This page is organized by energy source and includes impacts such as water usage, emissions, local pollution, and wildlife displacement.
Energy is sustainable if it "meets the needs of the present without compromising the ability of future generations to meet their own needs." Definitions of sustainable energy usually look at its effects on the environment, the economy, and society. These impacts range from greenhouse gas emissions and air pollution to energy poverty and toxic waste. Renewable energy sources such as wind, hydro, solar, and geothermal energy can cause environmental damage but are generally far more sustainable than fossil fuel sources.
The Diablo Canyon Power Plant is a nuclear power plant near Avila Beach in San Luis Obispo County, California. Following the permanent shutdown of the San Onofre Nuclear Generating Station in 2013, Diablo Canyon is now the only operational nuclear plant in California, as well as the state's largest single power station. It was the subject of controversy and protests during its construction, with nearly two thousand civil disobedience arrests in a two-week period in 1981.
A fossil fuel power station is a thermal power station which burns a fossil fuel, such as coal, oil, or natural gas, to produce electricity. Fossil fuel power stations have machinery to convert the heat energy of combustion into mechanical energy, which then operates an electrical generator. The prime mover may be a steam turbine, a gas turbine or, in small plants, a reciprocating gas engine. All plants use the energy extracted from the expansion of a hot gas, either steam or combustion gases. Although different energy conversion methods exist, all thermal power station conversion methods have their efficiency limited by the Carnot efficiency and therefore produce waste heat.
Climate change mitigation (or decarbonisation) is action to limit the greenhouse gases in the atmosphere that cause climate change. Climate change mitigation actions include conserving energy and replacing fossil fuels with clean energy sources. Secondary mitigation strategies include changes to land use and removing carbon dioxide (CO2) from the atmosphere. Current climate change mitigation policies are insufficient as they would still result in global warming of about 2.7 °C by 2100, significantly above the 2015 Paris Agreement's goal of limiting global warming to below 2 °C.
Total energy consumption in the United Kingdom was 142.0 million tonnes of oil equivalent in 2019. In 2014, the UK had an energy consumption per capita of 2.78 tonnes of oil equivalent compared to a world average of 1.92 tonnes of oil equivalent. Demand for electricity in 2023 was 29.6 GW on average, supplied through 235 TWh of UK-based generation and 24 TWh of energy imports.
A low-carbon economy (LCE) is an economy which absorbs as much greenhouse gas as it emits. Greenhouse gas (GHG) emissions due to human activity are the dominant cause of observed climate change since the mid-20th century. There are many proven approaches for moving to a low-carbon economy, such as encouraging renewable energy transition, energy conservation, and electrification of transportation. An example are zero-carbon cities.
Energy planning has a number of different meanings, but the most common meaning of the term is the process of developing long-range policies to help guide the future of a local, national, regional or even the global energy system. Energy planning is often conducted within governmental organizations but may also be carried out by large energy companies such as electric utilities or oil and gas producers. These oil and gas producers release greenhouse gas emissions. Energy planning may be carried out with input from different stakeholders drawn from government agencies, local utilities, academia and other interest groups.
Greenhouse gas (GHG) emissions from human activities intensify the greenhouse effect. This contributes to climate change. Carbon dioxide, from burning fossil fuels such as coal, oil, and natural gas, is the main cause of climate change. The largest annual emissions are from China followed by the United States. The United States has higher emissions per capita. The main producers fueling the emissions globally are large oil and gas companies. Emissions from human activities have increased atmospheric carbon dioxide by about 50% over pre-industrial levels. The growing levels of emissions have varied, but have been consistent among all greenhouse gases. Emissions in the 2010s averaged 56 billion tons a year, higher than any decade before. Total cumulative emissions from 1870 to 2022 were 703 GtC, of which 484±20 GtC from fossil fuels and industry, and 219±60 GtC from land use change. Land-use change, such as deforestation, caused about 31% of cumulative emissions over 1870–2022, coal 32%, oil 24%, and gas 10%.
Low-carbon electricity or low-carbon power is electricity produced with substantially lower greenhouse gas emissions over the entire lifecycle than power generation using fossil fuels. The energy transition to low-carbon power is one of the most important actions required to limit climate change.
Energy in Switzerland is transitioning towards sustainability, targeting net zero emissions by 2050 and a 50% reduction in greenhouse gas emissions by 2030.
The environmental impact of the energy industry is significant, as energy and natural resource consumption are closely related. Producing, transporting, or consuming energy all have an environmental impact. Energy has been harnessed by human beings for millennia. Initially it was with the use of fire for light, heat, cooking and for safety, and its use can be traced back at least 1.9 million years. In recent years there has been a trend towards the increased commercialization of various renewable energy sources. Scientific consensus on some of the main human activities that contribute to global warming are considered to be increasing concentrations of greenhouse gases, causing a warming effect, global changes to land surface, such as deforestation, for a warming effect, increasing concentrations of aerosols, mainly for a cooling effect.
Greenhouse gas emissions are one of the environmental impacts of electricity generation. Measurement of life-cycle greenhouse gas emissions involves calculating the global warming potential (GWP) of energy sources through life-cycle assessment. These are usually sources of only electrical energy but sometimes sources of heat are evaluated. The findings are presented in units of global warming potential per unit of electrical energy generated by that source. The scale uses the global warming potential unit, the carbon dioxide equivalent, and the unit of electrical energy, the kilowatt hour (kWh). The goal of such assessments is to cover the full life of the source, from material and fuel mining through construction to operation and waste management.
Energy in Sweden is characterized by relatively high per capita production and consumption, and a reliance on imports for fossil fuel supplies.
The Energiewende is the ongoing energy transition by Germany. The new system intends to rely heavily on renewable energy, energy efficiency, and energy demand management.
Environmental issues in the United Arab Emirates (UAE) are caused by the exploitation of natural resources, rapid population growth, and high energy demand. The continuing temperature rise caused by global warming contributes to UAE's water scarcity, drought, rising sea level, and aridity. The UAE has a hot desert climate, which is very vulnerable to the effects of climate change and contributes to worsening water scarcity, quality, and water contamination.
World energy supply and consumption refers to the global supply of energy resources and its consumption. The system of global energy supply consists of the energy development, refinement, and trade of energy. Energy supplies may exist in various forms such as raw resources or more processed and refined forms of energy. The raw energy resources include for example coal, unprocessed oil & gas, uranium. In comparison, the refined forms of energy include for example refined oil that becomes fuel and electricity. Energy resources may be used in various different ways, depending on the specific resource, and intended end use. Energy production and consumption play a significant role in the global economy. It is needed in industry and global transportation. The total energy supply chain, from production to final consumption, involves many activities that cause a loss of useful energy.
The environmental impact of bitcoin is significant. Bitcoin mining, the process by which bitcoins are created and transactions are finalized, is energy-consuming and results in carbon emissions, as about half of the electricity used in 2021 was generated through fossil fuels. Moreover, bitcoins are mined on specialized computer hardware with a short lifespan, resulting in electronic waste. The amount of e-waste generated by bitcoin mining is comparable to that generated by the Netherlands. Scholars argue that bitcoin mining could support renewable energy development by utilizing surplus electricity from wind and solar. Bitcoin's environmental impact has attracted the attention of regulators, leading to incentives or restrictions in various jurisdictions.
The environmental impact of Big Tech is a phenomenon in which many aspects of Big Tech contribute to negative impacts on the environment and climate change. In the big data age, technologists and people in general find it valuable to view emerging technologies with a critical lens, one of which is geared toward the environment. As these emerging technologies become more popular, they consider the extent at which they contribute to changes in the environment and whether they are inherently positive or negative.