Can Machine Learning and AI Help Restore Coastal Ecosystems?

By Kayla Perry '25

---

The increasing use of artificial intelligence (AI) and machine learning is one of the most recent advances in technology that has already begun to drastically change our society, learning techniques, and workplace environments. Artificial intelligence is currently being used in the field of environmental restoration, especially in coastal regions where the balance of ecosystems is disrupted by human activity. One prime example is Costa Rica’s “De Vuelta a Casa” project, which used machine learning to classify and return seashells to their natural habitats. This article will examine the benefits of AI within coastal restoration, while analyzing the potential legal, ethical, and technological consequences involved when utilizing technology within natural environments.

The Effect of Tourism on Coastal Ecosystems

Coastal ecosystems have an intricate composition, considering their environment, organisms, and human activity. (1) The populations of organisms and the sustainability of their ecosystems are greatly impacted by biological interactions with their environment. Because these components are interconnected and dependent on each other, a single change in the smallest link can have a major positive or negative effect on the entire ecosystem as a whole. (2) A recent example of this phenomenon is the removal of seashells from coastal ecosystems. 

At first, the removal of seashells by tourists seems like it would have little to no impact on an entire ecosystem; however, as tourism has increased over the years, so has this extraction. When shells are removed from their native habitat, local marine ecosystems are disrupted, and beach erosion is accelerated. (3) Many animals depend on seashells as a means of habitat and protection, and therefore suffer as an unintended, latent function of a seemingly innocent activity. 

         Another research study, conducted by the Florida Museum at the University of Florida partnered with the University of Barcelona, concluded that increased tourism on the coast of Spain has led to a decrease in mollusk shells by 70% during summer months. (4)  Mollusk shells, along with many other seashells, are necessary sources of calcium carbonate for marine organisms, provide crucial habitats for marine life, help enhance sand stabilization, and improve water quality. (5) Their removal reduces vital sources of nutrients, alters sediment chemistry, and can ultimately lead to harmful impacts on our marine ecosystems. (6) This drastic decrease is yet another example of how the removal of seashells in mass amounts can lead to long-term environmental impacts, especially within coastal ecosystems. As tourism has continued to increase over the past decades, the legal structure and policy at which society addresses these issues must evolve as well, especially with the implementation of artificial intelligence into environmental conservation.

         Currently, there are existing legal protections in place to address issues regarding environmental conservation. Florida Statute 259.032, Conservation and Recreation Lands, states that the preservation of public ownership over Florida’s natural areas is essential for protecting environmental resources, maintaining air and water quality, supporting sustainable water systems, restoring degraded public lands, and ensuring access to natural spaces for recreation. (7) There are also broader policies in place regarding international environmental law, such as the United Nations Convention on the Law of the Sea, which provides a legislative framework for preserving and sustaining marine life and environments. (8) However, a main issue within each of these frameworks is that the use of artificial intelligence and machine learning has yet to be addressed, nor have there been any recent laws discussing these new regulatory implications of using AI as a means of environmental conservation.

Case Study: “De Vuelta a Casa” 

The "De Vuelta a Casa," or "Back Home," initiative in Costa Rica is one of the most current and creative uses of artificial intelligence in environmental restoration. The project was created by the company Cervecería Imperial who partnered with the Costa Rican Ministry of Environment and Energy. (9) Researchers used machine learning to classify different seashells that were taken by tourists and confiscated at the San José airport in Costa Rica. The goal of the initiative was to return the shells back to their native coastal habitat. This issue stemmed from scientists in the past struggling to identify which coast, the Caribbean Sea or the Pacific Ocean, the millions of shells collected at the airport belonged to. Not only was the process of identifying shells physically time-consuming, but there was also a risk of relocating the shells to the wrong ecosystem. This innovative AI program allowed scientists to identify, in seconds, more shells in mass amounts than ever before.

The AI model is programmed with thousands of images of native shells from both the Pacific and Caribbean coasts, determining the origin of a shell with over 90% accuracy. (10) To date, Imperial’s project has returned over 1.5 tons of shells, demonstrating how AI is already beginning to transform restorative environmental processes. (11) This new approach to environmental protection combines the most recent technology with policy and public awareness. However, there are still risks with any new technological innovation, raising critical questions about environmental governance and legal frameworks around natural materials. Policymakers must reform traditional legal principles regarding environmental protection and restoration, and adapt legal frameworks around the introduction of AI within our natural environments.

Legal and Policy Implications of AI-Driven Environmental Restoration

         The “De Vuelta a Casa” initiative demonstrates the clear legal gap within the rest of the environmental policy and technology field. There are laws in place to regulate the prevention of environmental harm, yet not for AI-led ecological restoration decisions. As AI is quickly becoming one of the most powerful tools in environmental protection, new legal and ethical considerations must be addressed. Artificial intelligence has already begun to prove its vast potential in catalyzing restoration processes and the environmental benefits that follow; however, it also raises questions about how far we can let machines make environmental decisions once reserved for humans.

         While Costa Rica’s project demonstrates the positive benefits of AI-driven restoration, its success also reveals an increasing issue in environmental jurisprudence. Statutes such as the Ley de Conservación de la Vida Silvestre (1992) were drafted in an era that could never have anticipated machine-led ecological decision-making. (12) Questions such as who is liable when an algorithm incorrectly classifies a species or disrupts an ecosystem during a purely automated restoration process must be considered. 

In the United States, policymakers are beginning to address similar gaps. The Artificial Intelligence Environmental Impacts Act of 2024 represents a preliminary step toward regulating how AI systems affect the environment. The act focuses on addressing not only the positive potential of the AI but also the negative consequences, such as power usage. (13) For instance, the bill would direct the Environmental Protection Agency to conduct a study of the environmental impacts of AI, including energy use, pollution, and electronic waste, requiring companies to disclose the environmental footprints of their AI systems. (14) This Act represents a necessary first step towards encouraging transparency within the realm of AI and environmental protection. 

Florida Property Rights and Responsibility in AI Restoration

When applying AI to Florida’s coastal environment, it’s critical that policymakers consider the potential risks of misclassification or other mistakes in machine learning and explicitly integrate this into the law. In Florida, existing liability standards, which primarily focus on human negligence, agency, or wildlife regulations, are not yet equipped to handle the complexity of artificial intelligence. Responsibility could fall on the state agency that supervised the project, the people who operated the technology, or the private company or individual that created the algorithm. Traditional liability doctrines such as negligence and agency law may prove to be insufficient when harm results from AI systems rather than a clear, identifiable human act. Responsibility may ultimately need to be shared across multiple parties involved, specifically where oversight failures or operational decisions collectively contribute to environmental damage. Without clear statutory guidance, uncertainty may discourage the use of AI in environmental and coastal restoration. 

This gap is further complicated by the ambiguous legal status of seashells under Florida law. Florida law employs a combination of wildlife protection rules, coastal management policies, and public trust principles when legally defining seashells. Collecting seashells is permitted in Florida; however, the legality of the action depends on whether or not the harvested seashell contains a living organism, the specific type of organism, and the location where it is collected. (15) Unlike in Florida, Costa Rica’s initiative operated within a clearly defined legal framework under its wildlife conservation laws, which treat seashells as protected ecological resources, allowing authorities to confiscate, classify, and return them through an established statutory process. (16) 

As Florida continues to face pressures from increased tourism, erosion, and wildlife habitat loss, it is essential that its legal framework develop clear liability guidelines before AI-based restoration tools are implemented on a larger scale. As technology continues to rapidly advance and open new doors for not only coastal restoration but also for multiple industries and domains, it is crucial that our legal frameworks continue to adapt and define the principles that will govern the use of these revolutionary emerging tools.

Endnotes

1. “Ecosystems,” NOAA Fisheries, National Oceanic and Atmospheric Administration, accessed December 1, 2025, https://www.fisheries.noaa.gov/topic/ecosystems.

2. “Coastal Ecosystems,” National Ocean Service, National Oceanic and Atmospheric Administration, accessed December 1, 2025, https://oceanservice.noaa.gov/ecosystems/coastalecosci/

3. “The Importance of Returning Seashells to Beaches for Wild Hermit Crab Conservation, ” Local Hermit Crab Observation Society, accessed December 1, 2025, https://lhcos.org/the-importance-of-returning-seashells-to-beaches-for-wild-hermit-crab-conservation/.

4. “Seashell Loss Due to Tourism May Have Global Impact,” Florida Museum of Natural History, accessed December 1, 2025,  https://www.floridamuseum.ufl.edu/science/seashell-loss-due-to-tourism-may-have-global-impact/.

5. Ibid., 4.

6. Ibid., 4. 

7. Ibid., 4.

8. Ibid., 4.

9. “Oceans and the Law of the Sea,” United Nations, accessed December 1, 2025, https://www.un.org/en/global-issues/oceans-and-the-law-of-the-sea,

10. “How AI Companies Are Protecting and Restoring Nature,” accessed December 1, 2025, https://www.weforum.org/stories/2025/10/ai-companies-protect-restore-nature/.

11. “Back Home / De Vuelta a Casa,” Imperial Beer, Imperial Costa Rica, accessed October 14, 2025, https://imperial.cr/devueltaacasa/en/

12. Ibid., 8.

13. Artificial Intelligence Environmental Impacts Act of 2024, S. 3732, 118th Cong. (2024). 

14. Ibid., 12.

15. Ley de Conservación de la Vida Silvestre, Ley No. 7317, Asamblea Legislativa de Costa Rica (1992).

16. “Of Course. Here Is the More Informative Article Written in English,” People of Costa Rica, accessed December 1, 2025, https://peopleofcostarica.com/2025/08/23/of-course-here-is-the-more-informative-article-written-in-english/.