Every agency leader I talk to is focused on what artificial intelligence (AI) can do for their mission. Far fewer are focused on what it takes to keep that AI running when the infrastructure underneath it comes under pressure – particularly from mounting supply chain cost increases – and that gap is exactly what should be worrying us. AI is becoming a genuine force multiplier for national security: sharper decision-making, faster threat detection, better-prepared responders. But none of that happens without trusted, timely access to data. And right now, the infrastructure that data depends on is under more strain than ever.
That strain shows up in how hard it has become to plan for, procure, and scale the infrastructure supporting that data. Geopolitical instability, volatile technology markets, and growing competition for AI infrastructure are putting pressure on the storage, compute, networking, and data-focused resources agencies need to support mission-critical workloads.
As agencies become more dependent on AI-driven operations, the ability to maintain resilient access to mission-critical data is becoming increasingly important to national security. Agencies need efficient use of existing infrastructure and flexible architectures to keep mission-critical data accessible, secure, and actionable even as infrastructure markets, software dependencies, and geopolitical conditions become more volatile.
Supply chain risk creates mission risk
With global demand for capacity far outstripping fabrication capabilities, the cost of the chips underpinning our compute infrastructure has skyrocketed, with some essential components seeing price increases of up to 10x in less than a year, according to a recent report. Government agencies have less flexibility in responding to supply chain shocks than the private sector does. Procurement cycles run for years. Budgets are set by annual appropriations. Acquisition rules leave little room to maneuver when costs spike or critical components become scarce. The same report predicts that by 2028, 60% of chief information officers who do not respond to rising infrastructure pricing changes will risk delivery failures and reputational damage because they will be unable to effectively deliver infrastructure services. That’s no distant risk. It’s a countdown.
Here’s what that looks like in practice: a modernization effort gets pushed back because storage costs jumped and a planned hardware refresh gets delayed. Suddenly an agency is running a cyber incident response on infrastructure it already knew was inadequate. Logs take longer to pull. Threat intelligence sits in a queue. Sensor data and records that should inform real-time decisions arrive too late to matter. Every delay is a blind spot, and adversaries are patient enough to find them.
Although it looks and feels like a procurement problem, agencies must resist the urge to treat it as such – to try to ride out until prices come down. Instead, they should recognize it as a data resilience challenge that goes to the heart of mission continuity. The agencies that will hold up under pressure are the ones fortifying and maximizing the resilience of the data infrastructure they already have, not the ones betting that the next budget will arrive on time and solve the problem.
Optimizing Infrastructure Amidst Market Volatility
Managing modern supply chain risks requires agencies to move beyond treating infrastructure as a series of isolated capital procurements and focus on building resilient, modern data architectures designed to handle unpredictable demands. There are a few specific steps they can take.
First, in the face of supply chain constraints and rising chip costs, agencies should use this moment to build systems that are more resilient, efficient, and adaptable. Traditional ownership models expose agencies to price volatility, refresh risk, and overprovisioning at exactly the time when flexibility matters most. By investing in efficient architectures and considering economic models such as hardware-as-a-service or pooled infrastructure, agencies can shift risk away from their balance sheet, align spend more closely to actual demand, and maintain continuity even when component markets are unstable. The result is not just better cost control, but a stronger operational foundation that can scale with business needs.
Second, agencies can use existing resources more efficiently. Capacity constraints usually stem from duplicated data across systems, poor utilization, and workloads tied to specific systems rather than a lack of infrastructure. Reducing fragmentation and identifying inactive or redundant data and applying data reduction, data tiering, and lifecycle management can unlock capacity, improve performance, and extend the value of existing assets.
And finally, agencies should stop locking themselves into single vendors, platforms, or stacks and expecting flexibility to show up later. It won’t. Architectures built for portability – where data and workloads can move as mission needs shift – are what let agencies reduce duplication, simplify governance, and avoid becoming hostage to one supplier’s pricing or one platform’s roadmap. In this market, it’s the difference between adapting to disruption versus being defined by it.
Data resilience is mission resilience
Buying the right AI model is the easy part. What determines whether that investment actually pays off is whether the data behind it stays accessible, secure, and usable when conditions get unpredictable. Agencies that treat data infrastructure as core mission architecture, not IT overhead, are the ones that will still be operating effectively when markets tighten and supply chains falter again. The rest will be explaining the gap after the fact.
In a world where technology supply chains and geopolitical conditions grow more unpredictable by the day, data resilience isn’t just a component of mission readiness – it is its foundation.
