AI Makes Strategic Inroads in Data Centres, but Trust Remains a Barrier
In a revealing snapshot of the data centre industry, Uptime Institute’s 15th Annual Global Data Centre Survey (April–May 2025) shows that AI adoption is accelerating, but not without reservations.
Operators are increasingly deploying AI for key tasks like cost and capacity planning, energy optimization, and predictive maintenance. Roughly one‑third already run AI-based training or inference, and many more are preparing to introduce those workloads in the coming months.
What’s driving adoption and what’s holding it back?
Across more than 800 operators and over 1,000 vendors or consultants, the top concerns include:
Rising costs (76%)
Forecasting capacity amid volatile AI growth (71%)
Energy efficiency and cooling challenges (67%)
Power availability, supply chain disruptions, and talent shortages (all above 60%).
AI helps address these pain points. Survey respondents ranked its greatest operational benefits as enhanced facility efficiency (58%), fewer human errors (51%), and higher staff productivity (48%).
Trust levels vary by task
Confidence in AI is highly situational. While 70–73% trust it for sensor analytics and predictive maintenance, fewer are willing to let AI control core infrastructure or make configuration changes (35% and 14%, respectively).
As Uptime’s analysts point out, the industry is entering a phase of “careful adoption, testing, and validation” where selectivity, more so than unbridled enthusiasm, governs AI deployment.
Why this matters for you
For IT leaders and infrastructure teams, the survey highlights a strategic imperative: carefully calibrate AI rollout. Focus on incremental deployment of trusted use cases and shore up confidence through validated pilots. As AI workloads intensify, aligning IT infrastructure planning with evolving demand and managing trust boundaries will be essential.
In short: AI is gaining traction in data centres, but leaders must navigate adoption mindfully, balancing the allure of efficiency gains with keen awareness of what tasks AI can and cannot reliably perform.
Unplanned downtime and ageing infrastructure can hurt your data centre performance and drain your budgets.
Keep your infrastructure efficient and accurate with Teceze’s end-to-end data centre solutions.
Explore our data centre services here.
Bio-Industrialist & Bio-Tech Futurist at BioHack Startup & Founder President/CEO of AI and ML Startup
6dRight - with AI, Confidence is highly situational!
Masters in Computer Applications/data analytics
1wFully agree