Introduction  

  What is already known about this topic?  

A substantial body of research consistently shows that all forms of online and distance education contribute to reduced carbon emissions.  

  What the report adds  

The report presents a case study demonstrating how research in this field is conducted, offering practical examples of assessing the environmental impact of online education.  

Additionally, it expands the conversation by highlighting the impact of AI on the environment.  

  Implications for practice or policy  

Institutions could support online learning initiatives as part of broader strategies to meet environmental targets and Sustainable Development Goals.  

Institutions should develop and implement policies that promote the adoption of environmentally sustainable technologies, including AI.  

The environmental impact of online teaching and learning has become a high priority in line with the United Nation’s Sustainable Development Goals (SDGs). Research in the field overwhelmingly demonstrates a reduction in carbon emissions across all forms of online and distance learning. While recent research continues to appear promising, there are several considerations to ensure that sustainability is possible in the long term. For example, the carbon emissions emitted from residential homes where students may undertake online study must be considered to accurately measure carbon footprint. Broader shifts in higher education (HE) have seen a focus on other sustainable measures, such as solar panels, smart grids and composting on campus. However, these reductions can also contribute to a more sustainable approach as they reduce on campus carbon emissions while students learn online. The case study considers the significant challenge of providing a fully online engineering degree and the ways technology, like a virtual laboratory, is utilised to overcome and adapt to student needs. This technology continues to be a driving force in future developments. Further research is needed to determine the carbon footprint in developing countries, as well as the impact a ‘blended’ learning approach has on emissions.  

Current Trends  

In 2015, 17 Sustainable Development Goals (SDGs) were created to achieve a more sustainable future for all. In particular, a focus on quality education and climate action was among these goals. Universities have begun to address these goals in an attempt to reduce their carbon footprint in higher education (Li et al., 2021; Perales Jarillo et al., 2019; Filimonau et al., 2020). The COVID-19 pandemic provided the opportunity to operate a ‘control’ and analyse the effects of a fully online cohort due to campus closures (Filimonau et al., 2020).  

Previous research has highlighted a range of evidence to suggest that online teaching and learning reduces carbon emissions (Li et al., 2020; Carr et al., 2019; Versteijlen et al., 2017) and can reduce air pollution (Filimonau et al., 2020). Specifically, studies found a significant difference between the online and face-to-face carbon footprint:  

• Distance learning courses involve 87% less energy and 85% lower CO 2 emissions than the full-time campus-based courses. Part-time campus HE courses reduce energy and CO 2 emissions by 65% and 61% respectively, compared to full-time campus courses (Roy et al., 2008).  
• Distance-based models reduced energy consumption by 88% and achieved significant carbon reductions of 83% when compared with campus-based models (Caird et al., 2015).  
• E-learning offers a 20% and 12% reduction compared to print-based distance learning, due to increased energy use in computing and printing (Roy et al., 2008). However, as technology has become more sustainable, these reductions may have changed in recent years.  

• Online students’ emissions were, on average, one-third of those of their on-campus counterparts (Harlow, 2016).  
• With significant changes, a hybrid approach can achieve 15% reduction in emissions (Gamba et al., 2021).  

The most significant contributors to these stark differences include travel (Caird et al., 2015; Harlow, 2016; Gamba et al., 2021; Versteijlen et al., 2017) and campus and residential energy consumption (Caird et al., 2015; Harlow, 2016). In addition, climate-related or other disasters may necessitate emergency and online education (Selwyn, 2021; Heller et al., 2022) as well as provide flexible options for students (Ivancheva and Swartz, 2020; Perales Jarillo et al., 2019).  

However, the increasing carbon intensity of households (Filimonau et al., 2020) limits the sustainability of online teaching. Research warns that it is problematic not to account for consumption generated at home (e.g. energy and food) or where students undertake their study ( Ørngreen et al., 2019; Filimonau et al., 2020). Other challenges include the substantial energy required to operate campuses, even in a reduced capacity (Filimonau et al., 2020). Reconsidering the type of campuses built and retrofitting campuses to adapt to a more sustainable approach would reduce these limitations. Finally, a shift to online-only learning has the potential to increase the digital divide (Ivancheva and Swartz, 2020), as well as create pedagogical concerns around student engagement ( Ørngreen et al., 2019; Versteijlen et al., 2017; Ivancheva and Swartz, 2020). However, these concerns are widespread across online learning, regardless of sustainability and the carbon footprint associated with HE institutions.  

Technologies and Practices  

Technologies and practices for sustainability have rarely targeted online learning. Instead, the focus has been on other sustainable measures, such as sustainable infrastructure, covering energy, water and waste, in particular (Osorio et al., 2022). For example, Universidad Pontificia Bolivariana (UPB) in Colombia invested USD$4.5 million to construct a five-story building to meet requirements associated with energy efficiency, renewable energy and reuse of rainwater (Osorio et al., 2022).  

Solar panels, optimising the use of a smart grid, and introducing a composting station were some actions taken to reduce the carbon footprint of on campus learning. However, these measures are only likely to reduce the overall carbon footprint if institutions implement a shift to online only teaching and learning or a blended approach (Osorio et al., 2022). Similarly, outside of a shift to fully online learning, practices that reduce the biggest contributors to carbon footprint of on campus students, such as travel mode to campus, can be considered. Using the Netherlands as an example, Versteijlen et al., (2017) highlight that while the most common mode of travel to campus is train, these trains run on renewable energy and have no carbon footprint. This is compared to a high dependence on car travel to Australian universities which have a high-level carbon footprint. Finally, for online or blended modes, several practices can be used to reduce the environmental impact ( Ørngreen et al., 2019; Contreras-Taica et al., 2022). For example, “…the proper use of email, [such as] managing through labels or folders, the efficient use of laptops, [such as] lowering the brightness and disconnecting all USB ports that are not used…” (Contreras-Taica et al., 2022, p. 277).  

Case Study  

Perales Jarillo et al., (2019) highlight the case study of the International University of La Rioja (UNIR). UNIR is an online university committed to reducing social inequality and carbon emissions. While primarily based in Spain, the education institution services students in more than 77 countries through their online focus. Types of study include law, economics, education, social sciences, health, and engineering. Perales Jarillo et al., (2019) contend that UNIR students do not generate an impact on the carbon footprint on campus. While previous studies have highlighted issues with carbon footprint in homes or other study locations, Caird et al.’s, (2015) SUSteach model considers these factors. Based on 83% of carbon footprint being saved in a purely online model, Perales Jarillo et al., (2019) suggest that UNIR saves 136,663.68 tonnes of CO 2 , across their 41,870 students.  

A fully online engineering degree is seen as a significant challenge, particularly as much of the work requires ‘hands on’ and laboratory work. At UNIR, students engage with a collaborative environment including simulation tools and remote laboratories in circuitry. Simulation software offers sustainability in the exclusion of chemical or hazardous waste, while remote laboratories allow students to operate real equipment located in the university as an alternative to a ‘hands on’ lab session. One example of this is a remote laboratory for measuring radioactivity. Integrating a range of tools to break down geographical barriers has led to the global reach of the university, while simultaneously reducing the carbon footprint compared to traditional, on-campus universities.  

Future Developments  

Some challenges still exist for researchers considering the carbon footprint of online learning. Firstly, tools used to calculate CO 2 savings have not been applied widely in the developing world (Carr et al., 2019). This has specific impacts, particularly if students are still required to travel or consume other utilities to engage in online study (Haseeb et al., 2022). Secondly, in a post-COVID world, although there has been a shift by many universities to a ‘blended’ mode, the environmental impact of this mode of delivery has not been the focus of significant research (Contreras-Taica et al., 2022). Some studies calculate a possible decrease in environmental impact, but this would require a significant change in both regulations and logistics (Gamba et al., 2021). Other studies, however, suggest that blended approaches that use both campus facilities and online modes may contribute more CO 2 than online or traditional, on-campus modes (Caird et al., 2015; Versteijlen et al., 2017; Perez-Lopez et al., 2021). These approaches rely on face-to-face utilities and additional online learning facilities, creating additional energy consumed by the ICT facilities needed for online courses (Versteijlen et al., 2017). However, an increase in technology may reduce the carbon footprint further through, for example, the use of devices with lower emissions, like laptops and tablets as opposed to desktop PCs (Caird et al., 2015).  

What is the environmental impact of AI? 

AI offers considerable benefits for learning and teaching, transforming educational experiences through personalised or adaptive learning, tutoring systems, and predictive analytics. For instance, AI-powered systems and tools can make it possible to adapt content to individual learners’ needs, support early identification of students ‘at risk’, and reduce administrative workload for educators (Singh & Hiran, 2022; Chasokela & Hlongwane, 2025). These tools also enhance accessibility through features like real-time transcription, language translation, and automated feedback, ultimately promoting efficient educational environments (Sain, 2024).    

The integration of AI into learning and teaching contexts offers significant educational benefits, however, it also carries notable environmental impacts, primarily due to the energy consumption associated with AI model development and deployment. Training large AI models, particularly large language models (LLMs), requires vast computational resources, often powered by energy-intensive data centres. Although the research on the environmental impact of AI, particularly within educational contexts, has not been explored widely, recent studies in the US (Mehta, 2024; Scisson, 2024; Young 2024), report that data centres associated with AI development and deployment are placing significant strain on national electricity grids. Coulson (2024) states that data centres and data transmission networks were responsible for approximately 1–1.5% of global electricity consumption in 2020. They also produced an estimated 330 million tons of carbon dioxide-equivalent emissions, contributing roughly 1% of all energy-related greenhouse gas emissions. In addition, substantial energy is also required to cool the servers in these facilities, draining local water resources and contributing to drought (Mehta, 2024). This highlights the tension between technological progress within educational settings and its environmental impact.   

On the other hand, some studies have highlighted that AI can support more sustainable educational practices if implemented thoughtfully. For example, virtual classrooms and AI-driven personalised learning can reduce the need for physical infrastructure and travel, thereby lowering emissions (Chasokela & Hlongwane, 2025). Additionally, AI can optimise energy use in educational institutions by managing classrooms or guiding resource allocation based on data and analytics (Iatrellis et al., 2024). To balance the benefits of AI in education with its environmental impact, it is essential for institutions to consider sustainable AI approaches, such as using more efficient algorithms (Wu et al., 2022), choosing lower-impact cloud services (Patel et al., 2025), and prioritising carbon-aware AI development strategies (Schwartz et al., 2020).   

As AI tools become increasingly embedded in educational settings, so is the demand for constant data processing. The energy-intensive processes behind operating AI models contribute to rising carbon emissions, challenging the sustainability goals of many institutions. Balancing innovation with environmental responsibility will be essential for institutions to ensure that the advantages of AI within education are achieved without compromising the environment.  

While technology will continue to play an important role in the reduction of emissions and more carbon-efficient universities, it is equally important to recognise the importance of providing universal access to digital tools and learning environments. Ensuring that digital tools, like virtual laboratories mentioned in the case study, are accessible to all students is key to the establishment of inclusive and universally accessible online learning environments (Versteijlen et al., 2017; Perales Jarillo et al., 2019).  

References  

Caird, S., Lane, A., Swithenby, E., Roy, R., & Potter, S. (2015). Design of higher education teaching models and carbon impacts. International Journal of Sustainability in Higher Education , 16 (1), 96-111. https://doi.org/10.1108/IJSHE-06-2013-0065   

Carr, A., Modesto, S., Balasubramanian, K., Ortlieb, K., & Lesperance, J. (2019). Delivery Mode and Learner Emissions: A Comparative Study from Botswana. In W. Leal Filho & S. L. Hemstock (Eds.), Climate Change and the Role of Education (pp. 107-127). Springer International Publishing. https://doi.org/10.1007/978-3-030-32898-6_7   

Chasokela , D., & Hlongwane, J. (2025). Leveraging AI and ICT for Greener Education in the Face of Climate Change in Smart Classrooms. Journal of Research in Mathematics, Science, and Technology Education, 2 (1), 36-45. https://doi.org/10.70232/jrmste.v2i1.24  

Contreras-Taica, A., Alvarez-Risco, A., Arias-Meza, M., Campos-Dávalos, N., Calle-Nole, M., Almanza-Cruz, C., . . . Del-Aguila-Arcentales, S. (2022). Virtual Education: Carbon Footprint and Circularity. In A. Alvarez-Risco, S. S. Muthu, & S. Del-Aguila-Arcentales (Eds.), Circular Economy: Impact on Carbon and Water Footprint (pp. 265-285). Springer Singapore. https://doi.org/10.1007/978-981-19-0549-0_13   

Coulson, A. (2024). The environmental impact of AI in the lab: a double-edged sword? BioTechniques , 76(8), 353–356. https://doi.org/10.1080/07366205.2024.2376459  

Filimonau, V., Archer, D., Bellamy, L., Smith, N., & Wintrip, R. (2021). The carbon footprint of a UK University during the COVID-19 lockdown. Science of The Total Environment , 756 , 143964. https://doi.org/https://doi.org/10.1016/j.scitotenv.2020.143964   

Gamba, A., Maldonado, D., Rowen, M., & Torio, H. (2021). The Effect of the COVID-19 Pandemic on Mobility-Related GHG Emissions of the University of Oldenburg and Proposals for Reductions. Sustainability , 13 (14). https://doi.org/10.3390/su13148103   

Harlow, S. (2016). The carbon-based environmental impact of teaching and learning at the University of Waikato. Sustainability Symposium: Research in Action from Campus to Community, Hamilton, New Zealand.  

Haseeb, M., Tahir, Z., Batool, S. A., Majeed, A., Ahmad, S. A., & Kanwal, S. (2022). The Carbon Footprint of a Public Sector University Before and During the COVID-19 Lockdown. Global NEST , 24 (1), 29-36. https://doi.org/https://doi.org/10.30955/gnj.004222   

Heller, R., Sun, Y., Guo, Z., & Malik, A. (2022). Impact on carbon emissions of online study for a cohort of overseas students: A retrospective cohort study. 10 (849). https://doi.org/10.12688/f1000research.55156.5   

Iatrellis , O., Samaras, N., Kokkinos, K., & Panagiotakopoulos , T. (2024). Leveraging generative AI for sustainable academic advising: Enhancing educational practices through AI-driven recommendations. Sustainability, 16 (17), 7829. https://doi.org/10.3390/su16177829    

Ivancheva, M., & Swartz, R. (2020, 19 May). Universities go online during the pandemic: who reaps the profits? Corona Times . https://www.coronatimes.net/universities-go-online-pandemic-profits/  

Li, H., Khattak, S. I., & Ahmad, M. (2021). Measuring the impact of higher education on environmental pollution: new evidence from thirty provinces in China. Environmental and Ecological Statistics , 28 (1), 187-217. https://doi.org/10.1007/s10651-020-00480-2   

Mehta, S., (2024). Unraveling the environmental impact of AI: Opportunities and challenges. In Sage Business Cases. SAGE Publications, Ltd., https://doi.org/10.4135/9781071964859  

Ørngreen , R., Gnaur, D., & Henningsen, B. (2019). Meeting Online to Reduce Carbon Emissions and to Emphasise Values in Life and at Work (pp. 453-460,X,XII). Kidmore End: Academic Conferences International Limited.  

Osorio, A. M., Úsuga , L. F., Vásquez, R. E., Nieto-Londoño, C., Rinaudo, M. E., Martínez, J. A., & Leal Filho, W. (2022). Towards Carbon Neutrality in Higher Education Institutions: Case of Two Private Universities in Colombia. Sustainability , 14 (3). https://doi.org/10.3390/su14031774   

Patel, P. G., Dua, I. K., & David, S. (2025). Sustainable cloud development: optimize cloud workloads for environmental impact in the GenAI era (First edition.). Packt Publishing Ltd.  

Perales Jarillo, M., Pedraza, L., Moreno Ger, P., & Bocos, E. (2019). Challenges of Online Higher Education in the Face of the Sustainability Objectives of the United Nations: Carbon Footprint, Accessibility and Social Inclusion. Sustainability , 11 (20). https://doi.org/10.3390/su11205580   

Perez-Lopez, J.-B., Orro, A., & Novales, M. (2021). Environmental Impact of Mobility in Higher-Education Institutions: The Case of the Ecological Footprint at the University of A Coruña (Spain). Sustainability , 13 (11). https://doi.org/10.3390/su13116190   

Roy, R., Potter, S., & Yarrow, K. (2008). Designing low carbon higher education systems. International Journal of Sustainability in Higher Education , 9 (2), 116-130. https://doi.org/10.1108/14676370810856279   

Schwartz, R., Dodge, J., Smith, N. A., & Etzioni, O. (2020). Green AI . Communications of the ACM, 63(12), 54–63. Schwartz, R., Dodge, J., Smith, N. A., & Etzioni, O. (2020). Green AI. Communications of the ACM, 63(12), 54–63.  

Scisson, P. (2024, March 19). A.I. Frenzy Complicates Efforts to Keep Power-Hungry Data Centers Green. NY Times. https://www.nytimes.com/2024/02/29/business/https://www.nytimes.com/2024/02/29/business/artificial-intelligence-data-centers-green-power.html  

Selwyn, N. (2021). Ed-Tech Within Limits: Anticipating educational technology in times of environmental crisis. E-Learning and Digital Media , 18 (5), 496-510. https://doi.org/10.1177/20427530211022951   

Versteijlen, M., Perez Salgado, F., Janssen Groesbeek, M., & Counotte, A. (2017). Pros and cons of online education as a measure to reduce carbon emissions in higher education in the Netherlands. Current Opinion in Environmental Sustainability , 28 , 80-89. https://doi.org/https://doi.org/10.1016/j.cosust.2017.09.004   

Young, J. (2024, March 24). How Can We Harness the Power of Nature to Combat Ckimate Change . Newsweek. https://www.newsweek.com/why-ai-so-thirsty-data-centers-use-massive-amounts-water-1882374