Five barriers keep the digital services sector from reaching net-zero emissions

The Information Technology-Enabled Services sector has long enjoyed a reputation as a light touch on the climate compared with manufacturing, construction or power generation. A new systematic review argues that this comfortable image is increasingly out of date. Cloud computing, data centres, artificial intelligence and globally distributed digital operations are driving up electricity demand across the sector, and a synthesis of 78 studies published between 2009 and 2025 identifies five recurring, interdependent barriers that stand between ITeS organisations and credible net-zero transitions.

The review, published in Discover Energy, screened 1,342 records down to 78 articles through a two-phase process in which two reviewers independently assessed titles, abstracts and full texts, with disputes settled by discussion and, where necessary, a third researcher. The search covered Scopus, Web of Science, IEEE Xplore, ScienceDirect and Google Scholar, was restricted to English-language publications and was last updated in 2025. Roughly one-third of the included studies offered primary quantitative empirical evidence, about half were qualitative, case-study or policy analyses, and the remainder were conceptual papers or grey literature. The authors weighted the first two tiers more heavily when identifying barriers and used the third mainly for context.

The five barriers emerged inductively from thematic synthesis rather than from a predefined framework. They are the high energy intensity of digital infrastructure, dependence on carbon-intensive electricity grids, the absence of standardised sector-specific carbon accounting and reporting, fragmented policy and regulatory incentives, and weak corporate commitment combined with poor stakeholder alignment. The authors stress that these are not independent problems but a reinforcing system: high energy use only becomes an emissions problem when paired with fossil-heavy grids, inadequate carbon accounting hides the feedback signals needed to sustain investment, weak policy removes external pressure, and lukewarm leadership undermines internal action on everything else.

Energy intensity is the most visible barrier. Data centres are estimated to account for around 1 to 1.5 percent of global electricity use, a share expected to rise substantially as digital transformation and artificial intelligence expand. Cooling alone can consume as much as 40 percent of a data centre’s electricity, and the sector’s ICT industry as a whole is estimated to be responsible for between 1.5 and 4 percent of global greenhouse gas emissions. The review also highlights a less obvious culprit: inefficiency. Many data centres run at server utilisation rates as low as 10 to 30 percent, and over-provisioning, idle or so-called zombie servers, outdated hardware and poorly optimised cooling all add avoidable demand. The authors argue that distinguishing this avoidable energy use from the minimum required to deliver a given service is essential, because eliminating inefficiency is the most effective short-term emissions lever, more sensible than restricting the growth of legitimate digital services.

But efficiency gains only go so far when the electricity itself is dirty. Many ITeS operations sit in regions where grids still depend heavily on coal, natural gas or oil, making Scope 2 emissions from purchased electricity difficult to cut. In North America and Europe, power purchase agreements, renewable energy certificates and on-site generation have become more accessible, yet in many emerging economies these options remain scarce or unaffordable. The intermittency of solar and wind also raises reliability concerns for infrastructure that must run around the clock, often forcing backup from fossil sources where storage and grid flexibility are lacking. The review adds a nuance that is often missing from the debate: hyperscalers such as Google, Microsoft and Amazon have become among the world’s largest corporate buyers of renewable energy and have helped bring new wind and solar projects online, meaning some large firms can actively shape grid decarbonisation while smaller, geographically constrained companies remain genuinely limited by local conditions.

Measurement is the third barrier, and the review frames it as structural rather than merely procedural. General frameworks such as the Greenhouse Gas Protocol and ISO 14064 provide economy-wide guidance, but the authors conclude that none are suited to the actual operations of the ITeS industry. Shared, multi-tenant cloud environments make it hard to assign emission responsibility to individual clients, and inconsistent choices about reporting boundaries and emission factors reduce comparability between organisations. Crucially, the review argues that even if existing standards were applied strictly and uniformly, the sector’s largest emissions would still not be fully captured, because these frameworks were not designed to handle virtualised, geographically distributed workloads or the industry’s extensive Scope 3 emissions spanning cloud services, semiconductor supply chains and digital service provision.

Those Scope 3 emissions receive particular attention. Embodied carbon in servers, networking hardware and end-user devices accrues before any data is processed. Semiconductor manufacturing is energy- and resource-intensive, requiring ultrapure water, complex global fabrication chains and process gases with high global warming potential. Telecommuting shifts energy use into home offices and residential networks that fall outside traditional operational boundaries, while rapid device refresh cycles generate downstream e-waste emissions. The review identifies hardware circularity, careful selection of cloud providers based on grid carbon intensity, and supplier engagement as the largest immediate reduction opportunities, since these categories generally account for the biggest share of the sector’s footprint.

Policy and governance barriers complete the picture. National and regional climate policies tend to concentrate on industries that emit directly through fuel combustion, leaving the indirect emissions of digital services poorly covered. Voluntary pledges, green certifications and market-based tools such as renewable energy certificates exist, but the review finds they lack the power and reach to drive large-scale change without additional government rules. The authors are careful not to claim that more regulation is automatically better: poorly designed interventions can raise compliance costs, divert resources from innovation and burden small and medium-sized enterprises disproportionately. They favour outcome-based, proportionally scaled regulation that combines clear disclosure rules, consistent carbon pricing and stable renewable procurement rules with flexibility as technology evolves.

On the organisational side, the review documents a persistent gap between what companies promise and what they deliver. Many ITeS organisations publicly declare net-zero goals and renewable electricity commitments, but these are frequently not integrated into IT operations, procurement, finance and supplier management. The disparity is reinforced by unclear executive accountability, low internal pricing of carbon-related risk, competing capital-allocation priorities and unpredictable payback periods for green IT investments. The authors note that unless stated targets are backed by measurable interim milestones, cross-functional responsibility, capital planning and transparent performance monitoring, they may not translate into proportionate reductions in Scope 2 and Scope 3 emissions.

The review also maps emerging pathways, including renewable electricity procurement, carbon intelligence platforms that use real-time sensor data and machine learning to track emissions, circular ICT practices such as equipment refurbishment and certified recycling, science-based targets, supplier engagement and climate-smart infrastructure. It reports that electricity demand from AI-optimised data centres could more than quadruple by 2030, and cautions that efficiency gains can trigger rebound effects while AI-based optimisation itself consumes energy. The authors acknowledge their limitations candidly: the evidence base is concentrated in North America, Europe, India and China, English-language and peer-reviewed sources may exclude relevant regional work, and the maturity model and prioritisation frameworks they propose are analytical summaries rather than empirically validated tools. The review does not produce primary emissions data or assess intervention effectiveness. Its central message is that decarbonising the digital services economy is a systemic challenge, not a collection of separate technical tasks, and that strategies developed in digitally mature economies will need adaptation, particularly in developing regions where grids are dirtier, green finance is scarcer and regulatory frameworks remain immature.

Subject of Research: Barriers to achieving net-zero emissions in the information technology-enabled services sector

Article Title: A critical review of barriers to achieving net-zero emissions in the information technology-enabled services sector

Article References: Jindal, S., Jain, H., & Singh, L. (2026). A critical review of barriers to achieving net-zero emissions in the information technology-enabled services sector. Discover Energy, 6(1), Article 85. https://doi.org/10.1007/s43937-026-00199-4

Image Credits: AI Generated

DOI: 10.1007/s43937-026-00199-4

Keywords: net-zero emissions, ITeS, data centres, cloud computing, Scope 3 emissions, carbon accounting, renewable electricity, systematic review, e-waste, semiconductor supply chain, climate policy, digital sustainability