In 2025, talk of the artificial intelligence (AI) bubble grew stronger, with frontier projects such as Stargate dominating the headlines.
Meanwhile, server densities continued to grow to satisfy the voracious demand for compute placed by AI workloads.
As a result, electrical grids are under unprecedented pressure, creating environmental and economic concerns.
The industry is building ahead of projected demand to secure capacity, and financial models are evolving, with private capital playing a growing role in the industry. These trends will accelerate in 2026, making for an intriguing year, according to GlobalData.
Beatriz Valle, senior analyst: Enterprise Technology & Services at GlobalData, comments: “The pipeline of large-scale data centre projects will continue to grow in 2026. GlobalData estimates that there is currently a global pipeline of large-scale data centre projects with a total value of $2 306-billion. Data centres have become the backbone of economic competitiveness and a focus of geopolitical interest, driven by the AI revolution that commenced in late 2022.”
GlobalData’s report, 2026 Enterprise Predictions: Data Centre reveals that the semiconductor space will continue to evolve, with new entrants in the market trying to capitalize on the growing inference opportunity as the technological landscape evolves and more enterprises deploy generative AI in production environments.
Greater adoption will also drive edge computing forward, as real-time workloads drive the increased need for localized data centre infrastructure to support latency-sensitive applications.
Valle says: “With the changes that will take place, innovation will continue apace. Liquid cooling technologies will evolve to help dissipate heat in increasingly dense environments. Modular, pre‑engineered AI systems that can be added to existing data centres with minimal disruption will be marketed to help meet demand for compute.
“AI workloads will grow their presence and expand from AI labs to enterprise environments, a trend that will continue to shape the data centre market.”
While there are benefits associated with all these advancements, there are downsides too. Greater financial discipline will be required to bridge the imbalance between AI revenues and data centre capex, and regulatory oversight will become increasingly necessary to control the unintended consequences of data centre expansion.
Valle concludes: “Across 2025, the growing importance of AI infrastructure has made the data centre sector not only a catalyst for economic growth, but also a source of complexity, surrounding environmental and geopolitical issues, and this will continue in 2026.”