As 2025 draws to a close, the information technology sector finds itself at a pivotal moment in its journey toward environmental accountability. This year has not been defined by groundbreaking technologies aimed at decarbonisation, but rather by a significant shift in how sustainability is perceived within the industry. IT sustainability has evolved from a strategic differentiator to a critical operational and regulatory requirement, driven largely by the rapid integration of artificial intelligence (AI) into business operations.
The transformation marks a notable change: IT sustainability is now closely tied to business continuity, geopolitical supply chain risks, and mandatory financial disclosures. While this evolution indicates progress, it is largely motivated by necessity, reflecting the realisation that environmental missteps can result in direct financial penalties and increased liability for corporate boards.
Shifting the Sustainability Dialogue
The most significant change in 2025 has been the elevation of sustainability discussions to the core risk management agenda of executive committees. This shift is not optional; it is a response to impending regulations and the understanding that failing to address environmental issues can lead to measurable financial repercussions. Last year, conversations centred on the nebulous benefits of positive public perception; this year, the focus has shifted toward compliance with frameworks such as CSDDD, CSRD, and the updated SBTi Net-Zero Standard V2.
These regulations compel executives to move beyond general narratives and focus on the detailed, auditable data related to every asset, vendor, and cloud service utilised. For Chief Information Officers (CIOs), this translates into two major areas of concern. First, energy efficiency has been reframed as an essential cost of doing business, especially critical given the volatility of global energy markets. Secondly, the energy demands associated with generative AI have sparked internal debates regarding responsible computing architecture. Leaders must now justify AI investments not just through traditional return on investment (ROI) metrics, but also through a new lens of “return on compute,” which incorporates carbon expenditures into total cost calculations.
Despite this heightened focus, actual progress remains complex. The IT sector often lacks the authority to implement changes across various internal silos, and the budgets required for transformative initiatives are frequently limited.
Progress Amid Challenges
Despite systemic inertia, 2025 has seen tangible advancements in several operational areas, providing a potential roadmap for future net-zero initiatives. For instance, the decoupling of cloud growth from carbon emissions has emerged as a focal point. Hyperscale cloud providers have made significant strides in renewable energy procurement, leading to increased enterprise engagement in optimising physical operations. The adoption of advanced liquid cooling technologies has accelerated, particularly in hyperscale environments, aiding the scalability of AI solutions. Enterprises are now optimising workloads for low-carbon regions and utilising serverless architectures, achieving a level of separation between cloud expansion and emissions growth.
Another notable trend is the maturation of the Managed Device-as-a-Service (MDaaS) model. This approach has evolved into an essential environmental strategy, where enterprises outsource the entire device lifecycle, ensuring refurbishment and effective reverse logistics. Successful companies are leveraging these arrangements to reintegrate assets into the value chain through certified refurbishment, significantly reducing electronic waste. However, the widespread adoption of MDaaS remains limited, and the verification of these circular supply chains is still lacking robust third-party oversight.
Perhaps the most promising development is the emergence of green software engineering (GSE). This year marked a shift in focus from hardware alone, with organisations beginning to measure the energy consumption of code. Optimising algorithms and refactoring applications have become strategies to reduce reliance on resource-heavy computing. A significant milestone was the release of the W3C Web Sustainability Guidelines (WSG) Draft Note, which outlines structured best practices for minimising the environmental footprint of web products and services.
Addressing Gaps in Sustainability Efforts
Despite these advancements, 2025 also exposed critical gaps that could impede progress towards net-zero goals. One major challenge remains the measurement and reduction of Scope 3 emissions, particularly those arising from purchased goods and asset end-of-life processes. Many enterprises continue to depend on aggregated, industry-average supplier data, which is neither auditable nor sufficient for compliance with emerging regulations. The absence of granular product carbon footprints (PCF) from vendors hampers efforts to accurately assess and disclose emissions. This issue is compounded by the complexities of global supply chains, where suppliers are often reluctant to share detailed data due to competitive concerns.
Another pressing issue is the energy demand associated with generative AI. While AI offers the potential for optimising sustainability efforts, the immediate energy requirements of Large Language Models (LLMs) create a significant “energy debt.” The rapid adoption of AI technologies, combined with the high-performance computing resources necessary for their operation, risks negating environmental gains achieved elsewhere. Enterprises face a governance challenge, deploying AI solutions without comprehensive policies addressing model selection, inference efficiency, or resource decommissioning. Most organisations remain focused on achieving initial ROI metrics, relegating energy efficiency to a secondary priority.
Looking ahead to 2026, the IT Sustainability Think Tank emphasises the need to transition from merely identifying issues to actively addressing the gaps that hinder progress. Prioritising accurate data collection, enforcing necessary discipline, and managing the growing energy demands of AI must become integral components of future business resilience. As 2025 demonstrated, sustainability has transitioned into a core responsibility for corporate boards. The upcoming year must focus on translating these early efforts into full accountability, moving beyond aspirational goals to concrete, verifiable actions.
