Generative AI, sovereignty requirements and tighter financial scrutiny are changing how enterprises plan and operate hybrid cloud environments. In a forecast for 2027, Dennis Smith, distinguished vice president analyst at Gartner, argues that these pressures are increasingly connected: choices about workload placement can affect costs, compliance, governance and infrastructure capacity at once.
For CIOs, the task is not simply to add AI services or move data between environments. Organizations must decide where workloads can run, how policies apply across locations and whether power, cooling and budgets can support growth. Smith identifies five trends that leaders should account for as they shape hybrid cloud strategies.
AI workloads add infrastructure and operating demands
Generative AI can require high-performance GPUs, advanced networking and substantial storage, particularly for training large language models. Inference, meanwhile, is spreading across public cloud, edge and on-premises systems. Organizations may distribute it to address latency, resilience or data residency needs, but that distribution makes workload orchestration, networking, security and lifecycle management more complex.
AI infrastructure also places greater demands on power and cooling than traditional environments. Smith recommends locating workloads according to their requirements and business value: centralized cloud environments may suit model training and compute-intensive inference, while edge or on-premises infrastructure can serve inference that depends on low latency, resilience or local data handling. CIOs should assess capacity, power and cooling before demand exceeds available resources.
Sovereignty becomes part of cloud selection
Regulatory requirements and geopolitical tensions are raising the importance of digital sovereignty. Organizations need to understand where data resides, who operates the services and how much technical control they retain. Reliance on hyperscalers can expose enterprises to changes in regulation, geopolitical disruption and restrictions on cross-border data flows.
Regional or sovereign cloud options may help address some of those concerns, but they can also constrain access to advanced AI services or broad global integration. Smith advises CIOs to define and audit data, operational and technical sovereignty requirements across their cloud environments. Vendor assessments, he says, should consider sovereignty, compliance, operational independence and technical transparency alongside cost and functionality.
AI complicates cloud cost management
Large-model training and high-volume inference can make cloud spending less predictable. Energy consumption adds another component to the total cost of ownership, while rapid AI adoption can outpace an organization’s ability to track whether spending is producing business value.
Smith calls for stronger financial accountability, including cross-functional FinOps teams and automated cost controls. These practices can improve spending visibility and clarify responsibility for resource use. Energy should also be treated as a core cost metric, rather than only as a sustainability issue. The aim is to ensure that investment in AI scales in line with measurable value.
Distributed environments challenge governance
Applications, models and data increasingly span public and private clouds, edge systems and sovereign environments. Each can bring its own policies, controls and regulatory obligations. When governance is handled in separate organizational or technical silos, consistent enforcement becomes harder and gaps can emerge in compliance and audit readiness.
Smith recommends consolidating policies across environments and establishing consistent controls for AI deployments. Governance also needs to be shared across technology, legal, risk and compliance teams. That coordination can help organizations maintain oversight as workloads and data move between locations, rather than treating each platform as an independent policy domain.
Uncertainty raises the bar for investment decisions
Economic uncertainty is leading some organizations to defer major technology investments and large-scale projects, even as recurring cloud expenses continue. CIOs are under pressure to show the value of existing budgets and prioritize spending with clear business outcomes.
Smith cautions against treating a slower investment period as a reason to pause management improvements. Organizations can use the time to address inefficiencies, strengthen governance and compliance, and build support for priority initiatives by explaining their business value. Scenario planning and risk assessments can also prepare cloud operations for continued uncertainty or a faster return to investment.
Taken together, the five trends point to a more connected set of decisions for hybrid cloud leaders. AI placement affects infrastructure needs and spending; sovereignty shapes provider choices; and distributed operations test governance. CIOs that address these factors together can improve readiness without assuming that one cloud model will fit every workload.