As the calendar turns to the final pages of 2025, the information technology sector is at a critical juncture regarding its environmental responsibilities. This year was not marked by technological breakthroughs in addressing the decarbonization challenge, but by the decisive transformation of sustainability from a strategic factor to an operational and regulatory imperative.
This transition has required a painful reckoning with data complexity, supply chain realities, and the sheer energy appetite of modern computing, driven primarily by the rapid expansion of artificial intelligence (AI).
We entered 2025 with goals based on aspirations; we emerge under the obligatory mandate of reality. The major shift is profound: IT sustainability is no longer a parallel environmental, social and governance (ESG) initiative.
This has become closely intertwined with core business continuity, geopolitical supply chain risks and mandatory financial disclosure. While this shift represents progress, the impetus is driven more by necessity and the threat of liability than by shared ethical commitments.
The conversation moves from desirable to responsible
The most profound shift over the past year has been the forced inclusion of sustainability dialogue directly into the executive committee's core risk portfolio. This movement is not voluntary; it is driven by impending regulation and the sobering realization that environmental violations now carry direct, verifiable financial penalties and accountability at the board level.
Just a year ago, discussions revolved around the incalculable reputational benefits. Today, the lexicon is dominated by acronyms denoting mandatory compliance: KSDDD, CSRDand tightening SBTi Net-Zero Standard V2. These structures force executives to move away from narratives and toward detailed, auditable data associated with every asset, vendor, and cloud usage.
For the CIO, this manifests itself in two critical areas. First, energy efficiency is being decisively reimagined as a cost of doing business, critical to controlling operating costs in volatile global energy markets. Second, the sudden demand for energy from generative artificial intelligence has sparked intense internal debate about responsible computing architecture.
Leaders are increasingly being forced to justify investments in AI not only by traditional ROI, but also by an emerging “return on computing” model that necessarily integrates and accounts for carbon costs. This makes IT environmental costs an integral part of the total cost of ownership calculation rather than a polite footnote.
Despite this high-level engagement, progress remains challenging. IT often lacks the authority to drive change across complex internal structures, and the necessary budget and risk tolerance for truly transformative shifts remains extremely limited.
Real progress is when green shoots appear
Despite systemic inertia, 2025 brought solid, tangible progress in certain operational areas, offering a partial blueprint for future net-zero emissions efforts. Our confidence is supported by three examples, although it is important to recognize that large-scale implementation in the average enterprise is still in its infancy and is often limited to pilot programs:
1. Separating cloud growth from carbon: Hyperscale cloud providers have largely won the renewable energy procurement battle. The next frontier—optimizing physical operations—has seen the participation of enterprises. We have seen accelerated adoption of advanced liquid cooling technologies (still focused primarily in hyperscale environments, but critical to the future of AI scaling). Enterprises optimizing workloads for low-carbon regions and leveraging serverless architectures have successfully decoupled the rapid expansion of cloud technologies from the proportional increase in emissions. This success is largely due to hyperscalers, and the enterprise optimization campaign remains ongoing.
2. Development of a cyclical IT model (as a service): In 2025, Managed Device as a Service (MDaaS) has emerged as a critical tool for environmental protection. By outsourcing the entire lifecycle of a device, businesses are virtually committing themselves to remanufacturing and reliable reverse logistics. Successful businesses use these contracts to ensure assets are returned back into the value chain through certified repairs, dramatically reducing e-waste. The caveats are twofold: adoption of MDaaS is far from universal, and testing these round-robin chains still lacks the necessary and robust third-party oversight.
3. The Emerging Rise of Green Software Development: The formal emergence of green software engineering (GSE) is perhaps the most encouraging development. For too long, environmental considerations have been limited to hardware. This year, organizations began measuring code energy consumption – optimizing algorithms and refactoring applications to reduce dependence on resource-intensive computing.
An important event this year was the publication W3C Web Resilience Guidelines (WSG) Draft note. The guidelines, developed through a global collaborative effort in which I was pleased to participate, provide a structured and internationally relevant set of best practices for reducing the environmental impact of web products and services. Although the scope is focused specifically on the Internet rather than the entire spectrum of enterprise IT, the draft note nevertheless represents a significant step forward for the industry.
Persistent gaps undermine net zero momentum
For all the genuine acceleration, 2025 has been equally characterized by two persistent, critical gaps that threaten to derail the path to net zero and require urgent attention.
1. Scope 3 emissions gap: The most common and disappointing gap remains the measurement and significant reduction of Scope 3 emissions, especially from purchased goods and end-of-life assets.
Despite the urgency of regulation, the vast majority of businesses still rely on highly aggregated industry average supplier data (based on cost or activity) that is neither auditable nor sufficient for mandatory disclosure. The necessary mechanism—detailed product carbon footprint (PCF) data provided by each supplier—is simply not available at scale or with sufficient accuracy.
The challenge persists because it requires collaboration across complex, often private, global supply chains. Suppliers are reluctant to disclose detailed data, citing competition concerns, while buyers lack the leverage to demand it. The result is a “Scale 3 plateau”: targets are set, but baseline emissions remain stubbornly high, creating a significant confidence risk. We still primarily measure reflection rather than reality.
2. AI Generative Energy Debt: While artificial intelligence is a powerful tool for optimizing sustainable development, the immediate, unmanageable energy demand created by large language models (LLMs) represents a deep and growing gap. The speed of AI adoption, coupled with the need for expensive high-performance computing (HPC), is creating an “energy debt” that is erasing hard-won gains elsewhere.
The problem is management. Enterprises are deploying AI solutions without robust, enforceable policies for model selection, decommissioning efficiency, or decommissioning. It's important to note that most organizations remain focused on achieving initial ROI targets, relegating energy efficiency to optional performance tuning. Failure to enforce a “responsible computing” framework risks undermining the transformative power of AI due to its own growing environmental impact. This is the biggest risk in the IT sector's path to zero.
Strategic priorities for 2026 and beyond
As the IT Sustainability Think Tank looks to 2026, the focus must shift from identifying the problem to systematically closing remaining gaps in institutional discipline. We must view these priorities as non-negotiable elements of future business sustainability:
- Mandate details for area 3: Use purchasing influence to force suppliers to meet verifiable product carbon footprints (PCFs). The mandate must be non-negotiable, backed by clear supplier scorecards and contractual requirements.
- Institutionalization of green software development: Invest heavily in training and tools to integrate energy efficiency into the software development life cycle (SDLC). Software architecture must be treated with the same consideration for the environment as data center cooling, making efficiency a verifiable requirement.
- Manage AI energy costs: Implement a responsible AI framework that includes mandatory energy consumption metrics and resource allocation policies for all generative AI deployments.
In 2025, IT sustainability became the subject of a board audit. Next year should be the year when we finally collect detailed data, establish the necessary discipline and cope with the rapidly growing energy appetite of our own invention. The time for ambitious statements has finally passed; the immediate challenge now is to bring these nascent efforts to full and verifiable accountability.
