Sarah Fritz
Associate Editor
Loyola University Chicago School of Law, JD 2025
The use of artificial intelligence (AI) has been quickly evolving over the past few years, satisfying unexpected needs and allowing for innovative solutions to different problems. More recently, environmental organizations are increasingly interested in potential AI solutions for environmental issues.
AI has the potential to streamline regulatory compliance in the environmental context by switching the approach of environmental regulation from responding to current problems into prevention of future problems through data analysis, predictive analysis, and easier reporting procedures for violations of environmental regulation.
What is the impact of AI being able to record and analyze environmental data?
Climate change is an incredibly complex problem because the direct causes are not easily attributable to the environmental harm caused. However, AI can be utilized to collect and analyze data associated with climate change. AI would then be able to predict environmental degradation or harm caused by analyzing trends and patterns within available data.
AI can be extremely useful for analyzing the impacts of climate change because of its capacity to analyze large sets of data and create predictive models regarding the impacts of climate change in different areas. AI is able to evaluate the data sets presented and conclude regarding the potential impacts to the environment. Because of this ability, the use of AI in the environmental context would allow for more effective mitigation efforts.
The Environmental Protection Agency (EPA) currently uses predictive models of AI. The EPA’s Office of Compliance and the University of Chicago partnered to build an AI model aimed at promoting and increasing enforcement of environmental regulations. The predictive analysis resulted in a 47 percent improvement in detecting violations of the Resource Conservation and Recovery Act (RCRA), a statute related to federal regulation of toxic waste facilities. The EPA’s use of AI allows for prediction of environmental harm and promotes enforcement of environmental regulations.
AI will also be useful in collecting data associated with climate change. For example, the Arctic region is changing rapidly and warming at three times the global average. The significant warming of the Arctic region is causing sea caps and glaciers to melt as well as unknown effects on the marine and animal life in the region. During the spring, summer, and fall, ships in the Arctic region monitor the environmental impacts through collecting data. But, the ships do not monitor the environmental impacts or collect data during the winter because of ice conditions.
Therefore, AI can be extremely useful for regulatory bodies engaged in monitoring and analyzing climate change data in the Arctic when weather conditions do not permit traditional monitoring. The use of AI in this space will be incredibly useful as environmental degradation continues to alter our current climate. Proposals for AI monitoring of environmental data in the Arctic include the use of AI-powered robots. The robots would be utilized to monitor environmental conditions and provide data for regulators to utilize when making environmental legislation. In doing so, AI could be used to fill the gaps in data collection for environmental harms and allow for proper mitigation efforts.
What are the potential drawbacks of using AI in the environmental context?
While incredibly useful, AI has potential drawbacks for its use and can be subject to criticism in the environmental context. AI models can have negative environmental impacts, as they require vast amounts of energy and can cause large amounts of emissions. It is predicted that by 2025, the technology industry would consume up to 20 percent of the world’s energy and contribute up to 5.5 percent of global carbon emissions.
The training of AI models itself is very environmentally costly. The training of GPT-3, one model of AI, caused 552 metric tons of carbon emissions. For context, 552 metric tons of carbon emissions is equivalent to the emissions of 450 passenger vehicles. That is equivalent to 227,748 gallons of gasoline consumed by the general public.
It is hard to calculate the environmental impact of deployment of AI resources to the general public. Once released to the public, the environmental impact of AI use is hard to monitor because it is difficult to gauge the energy consumption of AI users. Due to this, many energy companies are introducing sustainability initiatives to offset the emissions caused through the use of AI by the general public.
The technology, energy and AI industries will have to weigh the benefits of using AI against the harmful environmental impacts that it can have. In a world that is continuously facing changes to the environment, the drive to create solutions to the climate crisis must be balanced against potential harms that could be caused by innovative solutions. Even though AI can be extremely useful in monitoring and analyzing the impact of climate change, that usefulness is deemed moot if the environmental impacts of AI itself outweigh the potential benefit it could have in analyzing environmental data.