Our industry is engaged in an important dialogue to improve the efficiency and resilience of real assets through transparency and industry collaboration. This article is a contribution to this larger conversation and does not necessarily reflect GRESB’s position.
Artificial intelligence is changing the data center sector from the inside out. The story is no longer just about demand growth; it is about whether the industry can expand capacity while managing the physical, environmental, and operational constraints that now define performance. Power availability, cooling intensity, water use, grid interaction, and transparency are becoming as important as location and connectivity. For GRESB readers, that makes data centers a clear example of how infrastructure quality is being redefined under pressure.
AI is accelerating an already complex market. Data centers have always been infrastructure-intensive assets, but the latest wave of demand is increasing the density of workloads and the strain on supporting systems. That means the most competitive facilities are no longer simply those that can house servers efficiently. They are the ones that can support high-density compute, maintain resilience, and show credible progress on sustainability and resource stewardship.
A sector under pressure
The first and most visible effect of AI is stress on the built environment. Higher-density workloads create stronger thermal loads, more demanding power profiles, and a greater need for infrastructure that can respond quickly and reliably. That changes the operating model for developers and owners. It is no longer enough to secure a customer and bring a building online; the site must be able to support changing workload profiles, evolving cooling requirements, and tighter alignment between IT and facility systems.
The challenge is not only technical. In many markets, the real constraint is the availability of usable power, the speed of grid connection, and the ability to deliver new capacity within a realistic development window. That has direct consequences for project timelines, leasing velocity, capital deployment, and long-term asset value.
Sustainability as a performance issue
Sustainability is no longer peripheral to data center performance. The sector’s growth has implications for electricity systems, carbon intensity, and in some cases water consumption, making resource management a central part of the investment case. GRESB’s Data Center Assessment reflects that shift. The purpose-built, sector-specific Assessment focuses on material issues such as power and grid interaction, community impact, energy and water intensity, and operational resilience, helping developers and operators communicate how they manage the sustainability risks and impacts most relevant to investors, utilities, regulators, and communities. The 2025 GlobalData report also highlights energy use, materiality, peer group selection, and data-sharing requirements.
That framing matters because it moves the discussion beyond generic ESG language. What counts now is whether growth is paired with better operational discipline. Can the operator show lower energy intensity over time? Can it align expansion with cleaner power procurement? Can it demonstrate responsible water management? Can it provide enough transparency for investors to compare performance across assets and peers? Those are the kinds of questions that matter to a GRESB audience because they relate directly to benchmarkable, comparable performance.
This is also where the sector’s long-term credibility will be tested. Demand growth is not enough on its own if the infrastructure behind that growth cannot keep pace with resource constraints. The strongest operators are those that can expand while improving visibility into energy use, asset health, and efficiency across the facility stack. In the end, that is the difference between growth that looks impressive and growth that is durable.
Colocation’s changing role
Colocation remains central to the data center ecosystem because many users want control, compliance, and network proximity without taking on the complexity of owning the full stack. In the AI era, that model becomes even more important. Enterprises still need flexible deployment options, low-latency connectivity, and access to secure environments that can support hybrid cloud architectures and evolving workload requirements.
For this reason, colocation is no longer just a space-and-power business. It is increasingly a service platform. Operators that can support high-density deployments, adapt cooling and energy systems, and integrate well with cloud and digital ecosystems are better positioned than those offering only commodity capacity. For investors and asset managers, that means the real question is not just how much capacity exists, but how capable the operating model is.
This is a useful lens for sector intelligence because it shows how the business model itself is changing. In an AI-driven market, the value of an asset depends on how well it supports performance, transparency, and adaptability.
Infrastructure is becoming intelligence
One of the strongest themes in the current market is that operational intelligence is becoming part of infrastructure itself. Data centers increasingly rely on digital systems that monitor, measure, and optimize energy use, cooling performance, and equipment health in real time. In a higher-density environment, that visibility is not optional. The cost of failure is too high, and the need for efficiency is too great. Tools for monitoring, energy management, and system integration are becoming essential enablers of resilient performance. The more complex the infrastructure becomes, the more valuable it is to have reliable, comparable data that can support decision-making across the portfolio.
The same is true for sustainability execution. A facility that can continuously monitor performance, identify inefficiencies, and respond to emerging risks is better equipped to improve over time. That is one reason the sector is moving toward integrated control environments rather than fragmented systems. As complexity rises, the ability to see and manage the whole asset becomes a competitive advantage.
What investors should watch
The next phase of the market will likely be shaped by three questions.
First, can the asset support AI-density workloads without compromising uptime or stability? This is the technical baseline for relevance. If a facility cannot reliably support the workload profile of the future, it will struggle to remain competitive.
Second, can the asset scale while staying within energy, water, and emissions boundaries? This is the sustainability test.
Third, can the operator turn complexity into a management advantage? This is where monitoring, analytics, and service capability matter. The most successful platforms will be those that make infrastructure easier to understand, easier to compare, and easier to improve over time.
These questions matter because they reflect a structural shift in the sector. AI is not only adding demand. It is changing the requirements for delivering that demand responsibly and profitably. The facilities that succeed will be those that combine physical resilience, sustainability discipline, and operational intelligence.
A more mature sector story
The AI era is forcing the data center sector to mature faster. In the past, growth could often be described in terms of expansion, connectivity, and adoption. Those factors still matter, but they are no longer enough. The market is now asking harder questions about how infrastructure is powered, cooled, monitored, and aligned with wider sustainability expectations.
AI is therefore not just transforming what data centers do. It is transforming how the sector is judged. The strongest assets will be those that can prove they are ready for denser workloads, cleaner growth, and more transparent performance. In that sense, sector intelligence is no longer just about reading the market. It is about understanding how the market itself is being redefined.
This article was written by Ekaterina Tsvetkova, Head of Advisory, at SE Advisory Services. Learn more about SE Advisory Services here.
Read more from our partners.