Balancing Innovation, Infrastructure & ROI
Artificial intelligence is commanding an extraordinary share of technology investment. According to Gartner’s September 2026 forecast, worldwide AI spending is expected to reach approximately $2.7 trillion in 2026, a 49.5% increase over the previous year. But as organizations accelerate their AI initiatives, another challenge is emerging: the infrastructure required to support those investments is competing for the same financial and operational resources.
A recent CIO report on AI spending and IT modernization highlights the growing tension. Citing research from Ensono, the article reports that 61% of surveyed organizations had paused, delayed, scaled back, or abandoned IT modernization initiatives during the preceding 24 months. Budget pressure, project complexity, and competing priorities, including AI, are contributing factors.
Meanwhile, expanding AI workloads are introducing new concerns around data center power availability, cooling capacity, hardware requirements, and operational costs.
For IT leaders preparing their 2027 strategies, the question is no longer simply how much to invest in AI… it is whether the organization can support the infrastructure required to make those investments successful.
The AI Spending Boom Is Creating an Infrastructure Funding Challenge
AI investments are accelerating across industries as organizations pursue opportunities in automation, analytics, research, and operational efficiency. The promise is significant, but realizing that value requires careful coordination between innovation and foundational IT investments.
When budgets are redirected toward new AI capabilities, organizations may postpone infrastructure upgrades, delay application modernization, or extend the life of systems that were already approaching their operational limits. These decisions can create unintended consequences.
AI initiatives depend on the very systems that modernization projects are designed to improve. Aging storage infrastructure, fragmented data environments, limited integration capabilities, and inconsistent governance can all restrict how effectively AI applications access and process information.
An organization may invest heavily in advanced computing resources only to discover that its underlying data architecture cannot deliver information efficiently enough to support the intended workloads.
The financial impact extends beyond delayed AI projects. Organizations may also face higher administrative costs, additional integration requirements, and infrastructure investments that must be revisited sooner than anticipated.
The solution is not to choose between AI and modernization… it is to recognize that modernization is often a prerequisite for AI success and establish a coordinated investment strategy that supports both objectives.
More Compute Does Not Automatically Mean AI Readiness
Much of the AI infrastructure conversation focuses on GPUs, accelerators, and specialized computing environments. These resources are essential for demanding workloads, but compute capacity represents only one component of a complete AI infrastructure strategy.
AI applications also require access to large volumes of data, often distributed across multiple systems, locations, and storage environments.
Without an appropriate data architecture, organizations risk creating bottlenecks that limit the value of their computing investments.
High-performance storage, scalable file and object architectures, efficient networking, and consistent data management practices are critical considerations for AI and high-performance computing environments.
The infrastructure must support how data moves through the complete workflow, from initial ingestion and preparation to processing, analysis, retention, and recovery. Not every dataset requires the same level of performance or availability.
Frequently accessed information may benefit from high-performance storage, while less active data can be managed through more economical storage tiers and archival solutions.
A comprehensive data lifecycle strategy helps organizations align infrastructure resources with actual workload requirements rather than continuously expanding expensive capacity.
Data protection must also remain part of the equation. AI environments may contain sensitive information, valuable research, proprietary datasets, and other assets that require appropriate governance and recovery capabilities.
For 2027, AI infrastructure planning should begin with an assessment of the entire data environment, not simply the acquisition of additional compute resources.
Power Is Becoming a Critical Infrastructure Constraint
The growth of AI is also changing the physical requirements of enterprise computing. High-density computing environments can place substantial demands on electrical systems, cooling infrastructure, and available data center capacity. These requirements are becoming increasingly important as organizations determine where and how to deploy new workloads.
According to the International Energy Agency’s Energy and AI report, global data center electricity consumption is projected to more than double to approximately 945 terawatt-hours by 2030, with AI representing the most significant driver of growth. The implications are particularly significant in the United States.
The U.S. Department of Energy reported that data centers accounted for approximately 4.4% of total U.S. electricity consumption in 2023. Its published projections estimate that share could increase to between 6.7% and 12% by 2028. These are projections rather than guaranteed outcomes, but they illustrate the scale of the infrastructure challenge.
For organizations planning AI expansion, power availability is no longer simply a facilities consideration. It can influence hardware selection, workload placement, data center location, operating expenses, and implementation timelines.
An organization may have the funding to acquire additional compute capacity but lack the electrical or cooling infrastructure required to operate it efficiently.
These dependencies need to be addressed early in the planning process, particularly for high-performance computing environments and large-scale AI deployments.
Where Will the Power Come From?
As AI infrastructure expands, energy sourcing is becoming part of the broader technology planning conversation. Grid capacity, utility interconnection timelines, generation availability, and the need for reliable electricity can influence where new data centers are developed and how existing facilities expand.
The IEA’s analysis of energy supply for AI identifies renewable energy and natural gas as major contributors to meeting additional data center electricity demand through 2030, with nuclear power expected to play a growing role toward the end of the decade and beyond.
Organizations are also exploring geothermal energy, on-site generation, battery storage, and other approaches to improve energy availability and operational resilience. However, these options present different considerations involving location, cost, reliability, development timelines, and infrastructure requirements.
The availability of generation capacity does not necessarily guarantee that electricity can be delivered to a specific facility when needed. For enterprise IT leaders, the practical takeaway is that infrastructure planning must account for both technological requirements and the physical environment supporting them.
Questions about available power, cooling capabilities, facility capacity, and future expansion should be addressed alongside decisions about computing and storage investments.
An AI strategy that ignores these dependencies may create financial commitments that are difficult to support operationally.
Infrastructure Efficiency Is Part of the AI ROI Equation
As infrastructure demands increase, efficiency becomes an important component of the business case for AI. Organizations cannot assume that expanding hardware capacity is always the most effective response to growing workloads.
A more deliberate strategy evaluates how existing resources are being utilized, where bottlenecks occur, and which investments can improve performance without introducing unnecessary complexity.
Enterprise storage and data lifecycle management have important roles in this process. Modern storage architectures can help organizations improve data accessibility, consolidate fragmented environments, and align performance with workload requirements. Intelligent data placement and tiering can also reduce reliance on premium infrastructure for information that does not require continuous high-performance access.
For organizations managing large research datasets, AI training environments, or high-performance computing workloads, these considerations can influence both infrastructure utilization and long-term operating costs.
Efficiency also extends to data movement. Repeatedly transferring large datasets between disconnected systems can consume network resources, increase processing delays, and introduce additional management requirements.
Architectures that simplify access to distributed information can help organizations reduce unnecessary movement while maintaining appropriate governance and security controls.
These improvements do not eliminate the power demands associated with AI. However, they can help organizations make more effective use of available infrastructure and avoid expansion driven by inefficient data management practices.
For Jeskell, this is where AI readiness connects directly to the broader data lifecycle. The objective is to ensure that every infrastructure investment contributes to performance, scalability, resilience, and long-term value.
Building a Smarter AI Infrastructure Strategy for 2027
The growing financial and physical demands of AI make coordinated planning increasingly important. Rather than evaluating AI initiatives as isolated projects, organizations should consider the infrastructure required to support them throughout their operational lifecycle.
That means understanding current capabilities, anticipated workload growth, data accessibility requirements, recovery objectives, and the true cost of operating the environment.
As IT leaders establish their 2027 priorities, several questions deserve attention:
- Can our existing infrastructure deliver the storage performance and data accessibility required by anticipated AI workloads?
- Are we funding the modernization initiatives necessary to support our AI investments?
- Have we accounted for power, cooling, facility capacity, and long-term operating expenses?
- Can we improve utilization of existing infrastructure before expanding capacity?
- Are our data protection and governance strategies keeping pace with changing workloads?
- Can we scale our environment without introducing unnecessary complexity or creating additional data silos?
Answering these questions provides a more complete understanding of AI readiness and creates opportunities to align technology investments with measurable operational objectives.
It also helps organizations identify where targeted modernization can support multiple priorities, improving data access, strengthening cyber resilience, and supporting AI initiatives without requiring an entirely new infrastructure environment.
Turning AI Ambition into Long-Term Value with Jeskell
AI has the potential to transform how organizations operate, but realizing that potential requires more than investment in emerging technology.
It requires an infrastructure foundation designed to support increasingly demanding workloads while balancing performance, security, scalability, and operational cost.
At Jeskell Systems, we help Federal and commercial organizations address these challenges through a comprehensive approach to data lifecycle management and infrastructure modernization. Our expertise includes enterprise storage, data protection, cyber resilience, high-performance computing, and the foundational infrastructure required to support AI and advanced analytics.
Through our extensive ecosystem of technology partners, we help organizations evaluate existing environments, identify infrastructure limitations, and develop integrated solutions aligned with their unique business and mission requirements.
With more than 35 years of experience, Jeskell understands that successful modernization is not about replacing technology for the sake of innovation. It is about maximizing existing investments while preparing for what comes next.
AI investments should accelerate organizational progress, not create new infrastructure limitations. As you prepare for 2027, let Jeskell help you build a scalable, resilient, and cost-conscious data infrastructure strategy that supports meaningful long-term value.