The future of AI depends on more than compute.
When reports surfaced that the U.S. Navy is evaluating whether the USS Gerald R. Ford could provide power to shore facilities while docked, most people viewed it as a military engineering story.
Technology leaders should pay closer attention. While the concept of a nuclear-powered aircraft carrier serving as a temporary power source may seem unrelated to enterprise IT, it highlights a challenge that is rapidly emerging across industries: the growing infrastructure demands of artificial intelligence.
For years, conversations about AI have focused on models, algorithms, and computing power. Yet as organizations move beyond experimentation and into production deployments, many are discovering that AI success depends on something much larger than the model itself… it depends on infrastructure.
AI Is Exposing Infrastructure Challenges That Have Been Building for Years
The race to adopt AI has accelerated investments in GPUs, high-performance computing, and advanced analytics platforms. However, these investments often reveal deeper challenges that have existed for years.
Data remains fragmented across multiple systems. Storage environments struggle to keep pace with growing data volumes. Governance policies are inconsistent across repositories. Network architectures were never designed to support large-scale AI workloads.
Then there is the issue few organizations expected to become a strategic concern: power.
Data centers supporting AI workloads consume significantly more energy than traditional enterprise applications. As demand for AI infrastructure grows, energy availability is becoming an increasingly important factor in infrastructure planning.
The Navy’s exploration of alternative power strategies is an extreme example, but it reflects a broader reality. Organizations everywhere are beginning to recognize that digital transformation initiatives are no longer limited by software alone. Physical infrastructure matters.
The Real AI Challenge Is Data Readiness
Many organizations approach AI as a technology initiative. The organizations seeing the greatest success increasingly treat it as a data initiative.
Artificial intelligence can only create value when it has access to trusted, governed, and accessible information. Yet many enterprises continue to operate with data spread across disconnected environments, legacy platforms, cloud services, and departmental silos.
The result is a common pattern: organizations invest heavily in AI technologies only to discover that their data foundation is not prepared to support them.
This is why data lifecycle management, governance, and storage architecture are becoming central components of AI strategy. Before organizations can scale AI effectively, they must ensure that data can be discovered, accessed, protected, and moved efficiently across the enterprise.
In many cases, infrastructure readiness becomes a greater challenge than model selection.
Resilience Is Becoming More Than a Cybersecurity Discussion
Cyber resilience remains a critical priority for every organization. However, resilience itself is evolving.
Today, resilience includes the ability to maintain operations despite disruptions to data, networks, supply chains, facilities, and even energy resources. As AI becomes more deeply integrated into business operations, infrastructure dependencies become more visible and more consequential.
Organizations may have robust cybersecurity programs, immutable backups, and disaster recovery plans. Yet a prolonged infrastructure disruption can still impact operations if the underlying systems supporting critical workloads are unavailable.
The lesson is clear: resilience must be evaluated across the entire infrastructure ecosystem, not just through the lens of cybersecurity.
A Systems-Level Approach to AI Infrastructure
The significance of the Navy’s experiment isn’t that an aircraft carrier might one day power a military installation.
It’s that organizations everywhere are beginning to recognize a new reality: infrastructure can no longer be treated as an afterthought.
AI may be the catalyst, but the underlying challenge is much broader. Data, storage, networking, resilience, and energy are becoming increasingly interconnected. Decisions made in one area now directly influence outcomes in another.
The organizations that succeed over the next decade will be the ones that recognize these dependencies early and build infrastructure strategies that account for the entire ecosystem, not just the technology trend of the moment.
As AI continues to reshape industries, the most important infrastructure conversations may no longer be about servers or software alone. They will be about how organizations create resilient, scalable environments capable of supporting innovation for years to come.