What’s Causing Healthcare’s Cloud Storage Budget Crisis: 2026 Wasabi Global Cloud Storage Index

Healthcare runs on data it can't throw away. Patient records, medical images, and clinical archives are retained for years, sometimes decades, to satisfy clinical and regulatory requirements.

Cloud storage was supposed to make data retrieval and storage easier with economies of scale, compliance-grade retention, and relief from the cost and footprint of on-prem hardware.

For the most part, it delivered. But it also introduced something healthcare budgets can't easily absorb: unpredictability.

Independent analysts at Vanson Bourne surveyed 1,700 IT decision-makers worldwide for the 2026 Wasabi Global Cloud Storage Index, including 171 in healthcare across North America, EMEA, and APAC.

Their answers explain what’s driving healthcare’s cloud storage costs and point to four pressures IT and finance leaders should understand before renewing their cloud storage contracts.

Where does the storage bill actually go?

On average, 49% of a healthcare organization's cloud storage bill goes to fees rather than to actual capacity. The cost of the bytes you store is often matched by the charges for accessing them: egress and ingress, per-request API fees, data retrieval, cross-region replication, object tagging, and even the requests to delete data you no longer want.

For four consecutive years, the ratio of these charges to the actual storage cost has hovered around 48–49%, and it's not improving.

Healthcare is more vulnerable to fees than most industries. Medical imaging studies in DICOM format (Digital Imaging and Communications in Medicine), electronic health record and electronic medical record systems, and clinical video archives don't sit quietly in cold storage. They're accessed, queried, and moved constantly, whether by a radiologist pulling priors, an EHR serving a chart at the bedside, a compliance team running a HIPAA audit, or a data science group feeding a model. Every one of those is a billable event, firing thousands of times a day.

The more a health system uses its data the way modern medicine demands, the more it pays in fees that have nothing to do with how much data it actually keeps.

Survey respondents feel it directly. The deeper problem, as one CIO/CTO in the study put it, is "variable and often opaque cloud pricing models" that make operating costs hard to predict and harder to control. 

Why are healthcare orgs exceeding their cloud storage budget?

Last year, 62% of healthcare organizations exceeded their cloud storage budget. That alone would be a planning problem. What makes it a structural one is the cause: 93% of the organizations that overspent pointed to higher-than-expected fees, not capacity.

When respondents named the specific culprits, two rose to the top: data operation fees (42%) and API call fees (37%). Both scale with activity, not storage volume, which is exactly why they're so hard to forecast. A hospital can predict how many terabytes it will store next year with reasonable confidence. Predicting how many times its systems will read, write, replicate, and retrieve that data across imaging, EHR access, backups, and disaster recovery drills is closer to guessing how many patients will come in on a random Tuesday in flu season.

Hyperscaler pricing turns that guess into a line item. When a bill carries dozens of fee types triggered by routine operations, accurate prediction becomes structurally impossible.

The budget alignment data bears that out. Sixty-two percent of organizations exceeded their budget, with the minority landing on target or under.

How is the market responding to unpredictable storage costs?

Last year, 73% of healthcare respondents used more than one public cloud provider for object storage. Multicloud isn’t the adventurous choice anymore, it’s the preferred path.

Application availability requirements (54%), wider performance options (53%), and better total cost of ownership (44%) are driving multicloud adoption. Spreading data across providers reduces dependence on any single vendor's pricing, performance, and uptime. For AI specifically, 65% of healthcare organizations use hybrid storage (mixing on-prem and cloud) to support their workflows.

In healthcare, hybrid is chosen for governance reasons as well as economics. Protected health information (PHI) carries data sovereignty and HIPAA obligations that don't always permit regulated patient data to move freely into public clouds. Hybrid lets organizations keep sensitive data on-prem when compliance requires it, while still tapping into cloud elasticity for the compute-intensive parts of an AI pipeline.

It isn't friction-free. As one healthcare CIO in the United States observed, running multiple clouds means you "have to really work on the network side of things" to keep performance from degrading across environments. Multicloud reduces single-vendor risk, but it has to be implemented deliberately to deliver on that promise.

Why aren’t AI investments paying off yet?

Commitment to AI in healthcare is now close to universal: 99% of respondents have an active AI infrastructure budget, and 64% expect that budget to grow over the next 12 months.

The intent is not in question, but the returns are. Only 34% of healthcare organizations say their AI projects deliver positive returns today. A further 41% expect to get there within a year. That's real movement, but it also means the majority still invests ahead of proof, spending now on the expectation that value arrives later.

Where the money goes explains a lot. Healthcare organizations direct 67% of their AI budgets to infrastructure (data, storage, and compute) versus 33% to software and applications. How efficiently that infrastructure runs determines whether any of it pays off.

The fee problem and the ROI problem are the same problem on two different line items. Every dollar lost to an unexpected storage fee is a dollar that never reaches clinical AI.

What is dark data costing healthcare organizations?

Dark data is the information an organization collects but never actually uses: the logs, call recordings, chat and email archives, and clinical video that get stored, retained, and then forgotten. For years it was an audit footnote. AI turned it into a liability and an opportunity at the same time.

The scale in healthcare is striking. More than half of organizations estimate that between 25% and 74% of their stored capacity is dark data.

That's not a rounding error. It's the digital equivalent of a radiology backlog, full of decades of imaging, research, and patient records that no model can learn from because no one can see, catalog, or reach it efficiently.

The intent to fix it is nearly unanimous: 95% of healthcare respondents call addressing dark data a strategic priority, and 47% rank it a high priority specifically. But intent runs into the same wall as everything else: legacy systems that don't interoperate, data trapped in operational silos, and migration costs that make action expensive even when the will is there.

Clinical AI is only as good as the data it can actually access, and right now much of healthcare's most valuable data is sitting in the dark.

What does this means for healthcare IT and finance leaders?

Healthcare IT and finance leaders can’t change how the hyperscalers bill. They can change how ready they are for it. Four moves matter most:

• Audit the whole bill, not just the storage line. Before renewing any cloud storage contract, map every fee category (egress, API, retrieval, and replication) alongside capacity. Among organizations that overspent in 2025, 93% traced it to fees they hadn't fully accounted for.

• Treat multicloud as a HIPAA strategy, not just a cost strategy. Distributing workloads across providers reduces fee exposure and single-vendor dependency at once, which matters as much for PHI and regulated patient data as it does for the budget.

• Don't build AI on top of invisible data. Dark data governance (metadata management, data cataloging, and storage visibility) is a prerequisite for clinical AI returns, not a post-launch cleanup task.

• Set a realistic AI ROI timeline and start measuring now. The healthcare organizations that put measurement frameworks in place early will be best positioned as the technology matures and budgets face scrutiny.

What should healthcare IT and finance leaders do now?

The 2026 Wasabi Global Cloud Storage Index reflects a healthcare sector that is allocating ample budget to AI, distributing risk across multiple clouds, and finally treating dark data as an asset worth unlocking.

What hasn’t been solved is the underlying economics. Fees that were already straining healthcare budgets are now compounding against AI cost pressure, and simply putting data to work is pushing budgets over the line unexpectedly.

If healthcare organizations stop treating storage as a fixed cost of doing business and start planning with surgical precision, budgets can hold steady. Predictable and transparent costs are a much-needed cure because every dollar reclaimed from fees is a dollar that can go to care, to research, or to the AI meant to improve both.