Bill Text (PDF)

Washington (February 1, 2024) – Senator Edward J. Markey (D-Mass.), chair of the Senate Environment and Public Works Subcommittee on Clean Air, Climate, and Nuclear Safety, Senator Martin Heinrich (D-N.M.), founder and co-chair of the Senate AI Caucus and a member of the Senate Energy and Natural Resources Committee, Representative Anna Eshoo (CA-16), co-chair of the House AI Caucus, and Representative Don Beyer (VA-08), vice-chair of the House AI Caucus, today introduced the Artificial Intelligence Environmental Impacts Act of 2024. The legislation would direct the National Institute of Standards and Technology (NIST) to develop standards to measure and report the full range of artificial intelligence’s (AI) environmental impacts, as well as create a voluntary framework for AI developers to report environmental impacts. 

The legislation also requires an interagency study to investigate and measure both the positive and negative environmental impacts of AI. While researchers increasingly highlight that AI can help tackle environmental challenges, such as by accelerating clean energy innovation, providing better weather forecasts, and improving cooling efficiency, the rapid growth of AI also comes with environmental harms. For example, increasing AI use could contribute to data center electricity demand doubling by 2026, leading to more carbon emissions. Demand for water to cool data centers is already creating concerns about water supplies, and the chips needed to run AI software are contributing to a growing mountain of electronic waste. 

“There is a Dickensian quality to the use of AI when it comes to our environment: It can make our planet better, and it can make our planet worse,” said Senator Markey. “Our AI Environmental Impacts Act would set clear standards and voluntary reporting guidelines to measure AI’s impact on our environment. The development of the next generation of AI tools cannot come at the expense of the health of our planet. I thank Senator Heinrich, Representative Eshoo, and Representative Beyer, for their partnership in making sure that AI contributes to a more livable future for generations to come.”

“AI offers incredible possibilities for our country, but that comes with high environmental costs,” said Representative Eshoo. “The resources necessary to research and develop AI are intensive, and as AI systems grow in scale and become more widely used across various sectors of society, it's critical to understand the environmental impacts of AI development and use. I'm proud to introduce the Artificial Intelligence Environmental Impacts Act to conduct a comprehensive study on AI’s environmental impacts, identify standards needed to measure those impacts, and create a system for AI developers to report the full range of their environmental impacts.”

“Understanding the environmental impacts of this quickly growing technology is critical so that we can begin to address those impacts,” said Representative Beyer. “While recognizing the ways AI can help us decrease emissions in other sectors and develop innovative climate solutions, we need to ensure we are being responsible with the adverse impacts it may have on our environment now.”

A copy of the legislation can be found HERE.

Cosponsors in the Senate include Senators Ron Wyden (D-Ore.), Peter Welch (D-Vt.), Alex Padilla (D-Calif.), and Cory Booker (D-N.J.). 

Specifically, the Artificial Intelligence Environmental Impacts Act would:

  • Require a Study on the Environmental Impacts of AI: The Environmental Protection Agency (EPA) would conduct a comprehensive study on the environmental impact of AI within two years. The study would examine AI models and hardware’s lifecycle, including energy consumption, pollution, and e-waste, as well as assess the positive and negative environmental impacts of AI’s applications. 
  • Convene an AI Environmental Impacts Consortium: NIST would convene a consortium of stakeholders to identify measurement needs and standards for AI's environmental impacts. 
  • Create a Voluntary Reporting System: NIST would develop a system for entities developing or operating AI to voluntarily report the full range of AI’s environmental impacts. 
  • Direct a Report to Congress: Within four years, the EPA, the Department of Energy, and NIST would submit a joint report to Congress, detailing the consortium's findings and describing the voluntary reporting system, as well as providing recommendations for further legislative and executive action.

The legislation is endorsed by Hugging Face, Data and Society, Climate Change AI, Public Citizen, Sierra Club, Electronic Privacy Information Center, Greenpeace USA, Center for AI and Digital Policy, Friends of the Earth Action, Kairos Action, Eko, Accountable Tech, Encode Justice, Union of Concerned Scientists, Fidutam, Green Web Foundation, Sustainable Digital Infrastructure Alliance, and Access Now.

“Recent advances in AI have great potential to help us tackle big societal challenges,” said Emma Strubell, Assistant Professor at Carnegie Mellon University. “However, there is increasing evidence that the development and use of AI can have a negative impact on the environment due to unprecedented computational requirements. Our understanding of the true scale and scope of these impacts is still nascent, limited by access to the necessary data and standards for reporting. Senator Markey’s AI Environmental Impacts Act provides a roadmap for elucidating the complex relationship between AI and the environment, which will be critical to ensuring that AI technology is developed and deployed so as to have a net positive impact on our environment and society.”

“The environmental impacts of AI technologies are undeniable - from the energy to power model training, and deployment, to the water needed to cool data centers and the rare earth metals to manufacture the hardware,” said Dr. Sasha Luccioni, Climate Lead at Hugging Face. “This bill will help shed some much-needed light on the extent of these impacts and how they’re evolving over time.”

“Senator Markey and Senator Heinrich’s legislation will address a critical, yet underexplored aspect of AI – its environmental footprint,” said Dr. Roy Schwartz, Senior Lecturer at the Hebrew University of Jerusalem. “The bill will promote transparency regarding the environmental aspects of AI tools, allowing organizations to both report and, as a result, mitigate their negative environmental effects. This act could not have come at a better time, as these tools are becoming ubiquitous, requiring larger amounts of energy to build and deploy them, water for cooling their servers, and land to build data centers for them.”

“This legislation provides a crucial step towards aligning society’s use of AI with climate change goals, by helping us measure the impacts of AI's inputs and applications, and therefore take more targeted action to shape those impacts,” said Dr. Priya Donti, Co-founder and Chair of Climate Change AI.

“As the use of AI rapidly increases, society needs to understand how this new technology impacts the environment,” said Erik Kojola, Senior Climate Research Specialist at Greenpeace USA. “This means we need a rigorous system for tracking energy use and carbon emissions. The Artificial Intelligence Environmental Impacts Act of 2024 is a vital step in creating that system and studying how AI contributes to the climate crisis. This will bring much-needed transparency and accountability to the emerging AI industry.”

“The Center for AI and Digital Policy strongly supports the Artificial Intelligence Environmental Impacts Act,” said Merve Hickok, President of the Center for AI and Digital Policy. “Senator Markey and Senator Heinrich’s bold initiative addresses one of the greatest challenges in the AI sector - the possible acceleration of climate change. We need research and analysis to assess the environmental impact of AI systems. The AI Environments Impacts Act should be a top priority for Senate consideration.”

“Artificial Intelligence has become the latest tool in the Silicon Valley corporate race, with every company trying to put out AI-driven products in order to stay competitive,” said Dom Leon-Davis, Deputy Director of Program and Strategy for Kairos Action. “If left unchecked, this rapid development pushed by tech executives can and will harm the planet we live on, particularly affecting poor, Black and Brown communities who already struggle to breathe clean air and drink clean water. From increased energy use to e-waste, and impacts that are unknown to us — we need regulation like this now. We’re glad to see this act take steps in the right direction of creating a world where tech works for all of us.” 

“The public deserves to know the very basics about how AI tools affect our material world,” said Caitriona Fitzgerald, Deputy Director at the Electronic Privacy Information Center.

“By creating transparency around AI’s impact on our environment, the Artificial Intelligence Environmental Impacts Act of 2024 will give the public and policymakers a tool for pressuring technology companies to develop their products in line with our collective health and safety."

“As Lord Kelvin wisely stated, ‘To measure is to know.’ We enthusiastically welcome and are eager to collaborate on Senators Markey and Heinrich’s initiative, as the United States takes a leading role in driving transparency and setting global standards,” said Max Schulze, Director of the Sustainable Digital Infrastructure Alliance. “This initiative crucially creates visibility for the environmental costs of artificial intelligence, encompassing not just the models but their usage as well. We particularly endorse the holistic approach taken, going beyond mere energy consumption to consider all environmental impacts across the AI supply chain. This is vital for developing sustainable and responsible AI practices globally.”

“Right now, there is significant uncertainty about the impact of AI on the environment,” said Peter Henderson, Assistant Professor at Princeton University. “On one hand, it could be used to improve infrastructure and systems for better efficiency, on the other hand it could harm the environment through increased demand for materials, manufacturing, and energy usage. The government has a large role to play in ensuring that AI is a net positive for the environment, by promoting transparency and innovation around AI’s environmental impacts. I applaud Senator Markey and Senator Heinrich for introducing this Bill that will provide a roadmap toward positive outcomes.”

“One of the many unknowns about AI is its impact on the environment,” said Kaili Lambe, Policy and Advocacy Director at Accountable Tech. “As rapid advances have led to increasingly widespread use of LLMs and other data-intensive AI systems, it’s imperative that we understand and take steps to mitigate unintended harm, including the potential for environmental degradation. Too often transparency takes a backseat when it comes to Big Tech innovation. Senators Markey and Heinrich’s legislation offers an opportunity to gain critical insight into the environmental impact of AI, and we are eager to see it pass.”

“AI technologies are spreading rapidly, with the potential for major impacts on society,” said Jennifer Jones, Director of the Center for Science and Democracy at the Union of Concerned Scientists. “We must fully understand both the benefits and the risks, of these technologies—including intensive energy consumption and creation of environmental waste that harms people and nature. We have a responsibility to track the environmental impacts of AI, to ensure this information is publicly available and to help us plan to mitigate those harms. Importantly, the bill will allow stakeholders, including civil society, to examine and report on the full life cycle of artificial intelligence models, systems, and hardware—bringing the voices of the most impacted communities to the table.”

“Right now, we have only a limited understanding of the depth and scale of AI’s environmental impacts—but what we do know is alarming,” said Brian J. Chen, Policy Director at Data & Society. “AI’s massive energy consumption and pollution effects are likely to exacerbate global warming and climate disaster. The AI Environmental Impacts Act lays an important research foundation for a better-grounded understanding of the technology’s harms to people, communities, and habitats.”

“What gets measured gets managed,” said Michelle Thorne, Director of Strategy and Partnerships at the Green Web Foundation. “We at the Green Web Foundation welcome this bill as an important step towards transparency of the environmental impacts of AI. To date, the lack of consistent and comparable data has been a blocker for technologists, researchers, policymakers and civil society. This bill will help enable more data-informed discussions about how resources are allocated to computing.”

“Friends of the Earth Action is proud to endorse the bill to address the climate concerns of artificial intelligence,” said Michael Khoo, Climate Disinformation Program Director at Friends of the Earth Action. “In addition to generating even more disinformation around the climate crisis, AI comes at an enormous energy cost that must be made transparent. Congress should not make the same mistake of simply trusting AI companies to do good, the way they did with social media over the last two decades. We call on lawmakers to support this bill for the sake of our planet.”