The opportunities side of the ledger is genuinely impressive. Digitalization—the process of converting physically collected information into computer-readable language—can overcome sustainability barriers that have frustrated policymakers for decades. Satellites and remote sensors now monitor ecosystem health in real time, giving conservationists an unprecedented window into deforestation, glacial retreat, and ocean warming. Artificial intelligence applications can improve efficiencies in agriculture and transport, reducing carbon and water footprints by optimizing irrigation schedules, logistics routes, and energy grids. Blockchain technologies can decentralize and expand access to energy and finance sectors, enabling peer-to-peer renewable energy trading and financial inclusion for communities excluded from traditional banking. Online platforms create new opportunities to access education, healthcare, and jobs, decoupling opportunity from geography. Taken together, these applications suggest that digital infrastructure could serve as the nervous system of a globally coordinated sustainability effort, detecting problems faster and allocating resources more precisely than any previous generation of tools allowed.

But the risks are equally concrete, and the authors document them with striking specificity. Data centers that train and deploy complex AI applications are highly energy- and water-intensive. One frequently cited example: a data center cluster in Iowa supporting the training of ChatGPT-4 used 6 percent of the district’s water during its last training month—a figure that surprises even technologically literate audiences, since the water footprint of computing is largely hidden from public view. Beyond operational resource use, digitalization drives demand for rare earth metals, whose extraction is linked to significant social and environmental harms, from toxic tailings to labor exploitation. The sector also generates staggering quantities of electronic waste: 62 billion kilograms in 2022 alone, according to the Global E-waste Monitor. Perhaps most insidiously, lower production costs may lead to higher production and consumption, a rebound effect that can entirely offset the efficiency gains that made the technology attractive in the first place. A cheaper, more efficient system that scales tenfold can consume more resources in absolute terms than the inefficient one it replaced.

The social risks compound the environmental ones. The digital divide—the gap between those with access to digital infrastructure and those without—continues to limit access to emerging online opportunities in education and employment for marginalized communities, meaning that digitalization can widen inequality even as it creates new forms of opportunity. Algorithmic bias and misinformation are also well documented, with biased training data producing discriminatory outcomes in hiring, lending, and policing, while generative systems accelerate the spread of false information at scale. The authors emphasize that the ubiquitous nature of digital technologies, their cross-boundary impacts, and their varying interactions under different socio-economic and political settings make it extremely challenging to map and measure the full spectrum of environmental and social consequences. A single AI model’s footprint spans mines on one continent, manufacturing on another, data centers on a third, and disposal sites on a fourth. Society, they argue, must proceed with caution and carefully manage the adoption, usage, and intended purpose of these technologies rather than assuming that innovation alone will bend the curve toward sustainability.

What distinguishes this piece from a general warning is its actionable structure. Drawing on their experience as convenors and knowledge creators at the sustainability-digital-society interface, and as co-champions of the Coalition for Digital Environmental Sustainability (CODES), the authors direct a set of priority recommendations at three influential stakeholder groups: sustainability science practitioners, including scientists, researchers, and community groups; the digital innovation sector, including technology developers, industries, and suppliers; and policy makers with regulatory influence. The premise is that no single group can manage digitalization’s footprint alone—collaboration across sectors, with each actor adopting priority actions within their own communities, is the mechanism by which sustainable digital practice becomes the norm rather than the exception.

For the research community, the authors argue that sustainability science practice can no longer be separate from digitalization, and they propose three concrete efforts to align the two. First, learning around digital tools, skills, risks, and opportunities should be integrated across sustainability education, so that established environmental studies curricula train future practitioners—including educators—to critically analyze and use digital technologies in a safe, sustainable, and effective way. Second, they call for building a sustainable digitalization community of practice, bringing together actors from diverse knowledge systems—industry, research, Indigenous, and policy—at the intersection of environmental sustainability and digital innovation. Such communities, they note, play a crucial role in solution co-design, knowledge sharing, and building the trust needed to overcome barriers to action. Third, they propose establishing a formal scientific process, analogous to existing climate assessment bodies, to systematically map the direct and indirect pathways through which digital technologies impact the environment. Such a process could identify data gaps, set standards for digital sustainability, identify key stakeholders and assign accountability, and establish science-based policies to govern digital technologies.

The digital innovation sector, meanwhile, is asked to accelerate efforts to monitor and mitigate environmental impacts across the entire technology lifecycle, with a particular emphasis on providing the data needed to measure and manage those impacts. The authors recommend that manufacturers of computational hardware report the emissions and resources extracted in the assembly and transportation of equipment, and that computer scientists report the resources used to train and deploy various AI applications. These life cycle analyses would provide a clearer understanding of how digital technologies affect the environment across supply chains and reveal opportunities to reduce impact—addressing precisely the opacity that currently obscures figures like the Iowa water consumption data. They also call on innovators to follow green digital infrastructure standards, adopting environmental requirements for the development and purchasing of digital technologies, such as minimum use of recycled materials, a requirement for data centers powered by renewable energy, and recycling of electronic waste.

Policy makers receive the third set of recommendations, centered on ensuring that digital solutions are designed and deployed with both people and the planet in mind. The authors urge decision makers to prioritize funding—including public digital infrastructure procurement—toward digital solutions that demonstrate measurable benefit to both people and the planet, rather than subsidizing growth for its own sake. They call for a strengthened regulatory landscape, with requirements at national, regional, and international levels to increase transparency on the environmental footprints of energy-intensive digital technologies such as large language models, build institutional capacity to recycle and reuse e-waste, and leverage low-carbon or renewable energy sources to power data centers. Finally, they recommend increasing public awareness by educating constituents on the environmental impacts of digital technologies—around e-waste, energy use, water consumption, and other issues—so that demand-side pressure reinforces rather than undermines sustainability goals.

The authors argue that governance momentum is already building, and they point to a rapidly maturing institutional landscape. The European AI Act is the first comprehensive regulation of artificial intelligence by a major regulator. The International Telecommunications Union offers guidance on assessing the environmental impact of the information and communication technology sector. The United Nations Global Digital Compact, adopted by world leaders at the UN Summit of the Future in September 2024, establishes shared principles for an open, free, and secure digital future. Globally recognized standards, such as ISO 42001 for the responsible development and use of AI systems, could be expanded to include environmental considerations. Yet the evidence also shows why targeted strategies remain essential: a recent analysis across European countries found a strong link between the digital economy and digital society, but this correlation did not strongly connect to improved environmental sustainability. In other words, digital development does not automatically become sustainable development—capacity building and awareness raising among technology developers, science funders, and the sustainability research community remain crucial steps toward that goal.

The stakes of getting this right could hardly be higher. Sustainable digitalization, the authors conclude, has the transformative potential to create a thriving future for people and the planet, but only if the environmental and social costs of the digital transition are governed with the same seriousness as its benefits are celebrated. The piece arrives at a moment when AI investment is surging globally and data center construction is accelerating, making the window for embedding sustainability requirements into infrastructure, curricula, and regulation both urgent and narrow. Whether the digital age becomes an era of planetary repair or one of accelerated extraction will depend, the authors suggest, on whether scientists, innovators, and policy makers act collaboratively—and quickly—on the priority actions they have laid out.

Subject of Research: Sustainable digitalization: governing the environmental and social impacts of digital technologies to advance global sustainability goals

Article Title: Striving for a sustainable planet in the digital age: Recommendations and priority actions

Article References: Kumarasinghe, N., Gleeson, E., & Ventimiglia, A. (2026). Striving for a sustainable planet in the digital age: Recommendations and priority actions. PLOS Sustainability and Transformation, 5(5), e0000249. https://doi.org/10.1371/journal.pstr.0000249

Image Credits: AI Generated

DOI: 10.1371/journal.pstr.0000249

Keywords: sustainable digitalization, artificial intelligence, data centers, e-waste, digital divide, climate change, sustainability governance, life cycle analysis, CODES, PLOS Sustainability and Transformation, Global Digital Compact, environmental policy