The Economics of Data: Scaling Infrastructure for the AI-Driven Data Economy

The impact is significant and accelerating. According to the IDC research, 94.7% of surveyed organizations are storing more data as a direct result of AI and generative AI adoption over the past 12 months. Sixty-one percent experienced data growth of 25% or more over the past year due to AI, and 74% expect that pace to continue, with data volumes growing 25% or more over the next three years.^*

Among surveyed organizations, 75.9% report bringing increasing volumes of archived cold-tier data back online to support AI workloads; 96% anticipate needing faster archive retrieval to support AI inference and retrieval-augmented generation (RAG) applications, and 98.2% consider total cost of ownership per terabyte as an important or very important factor in storage decision-making.^*

AI both draws on existing data and generates new data throughout its operational lifecycle. Model outputs, inference logs, and synthetic data are all byproducts of AI workloads that can persist long after the compute cycles that produced them. That data may then become an input for future AI applications, creating what we call a compounding data cycle: AI requires data, generates more of it, and increases the potential value of information organizations already hold.

It means that the traditional tiering model—where data is classified as either active or archived based on access frequency—is becoming less clear-cut. Information that once sat dormant in cold storage may now need to be retrieved quickly to support an AI inference request, a RAG query, or a workload that didn't exist when the data was originally stored. For infrastructure leaders, this raises the stakes for how data is retained and made accessible across its full lifecycle.

There is no single answer, but the IDC research points to a consistent set of priorities. Organizations need to evaluate storage across the full data lifecycle: considering what data needs to be immediately accessible, what can reside in other storage tiers, how quickly information may need to move between tiers, and what it costs to retain growing volumes of data over time. The goal is an infrastructure strategy that aligns resources with workload requirements while positioning organizations to extract long-term value from their data.