Cloud budgets hit 26% of IT spend, and the hiring plan is shifting to platforms

Cloud spending continues to climb. What is changing is where leaders are assigning the teams that make that spend deliver value.

CIO reported in late August that IT leaders expect 26% of their IT budgets will be allocated to cloud computing within the next year, citing the 2026 Foundry Cloud Computing Study. In that same Foundry survey, 74% of IT leaders said they accelerated cloud migrations in the last 12 months, up from 70% in 2025 and 63% in 2024, according to CIO.

Those are big-picture indicators. The operational signal, seen in hiring data and recent public-sector org charts, is similar: cloud teams are adding platform, governance, and adoption capacity, because those functions connect migration speed to day-two operations and business results.

Spending is climbing, but the challenge is making cloud run like a product

CIO’s coverage of Foundry’s survey also points to how teams justify cloud internally. Nearly three in four respondents (73%) said cloud capabilities helped their organizations generate higher, durable revenue over the past 12 months. The figure is self-reported by IT leaders, not a controlled causality study, but it matters because it is the narrative many CIOs take into budget discussions.

The same survey links cloud strategy and AI adoption. CIO’s reporting of Foundry’s data says 80% of respondents in North America and APAC reported that their cloud strategies accelerated AI adoption, compared with 68% in EMEA. For operators, the regional difference is a reminder that an “AI-ready cloud” is not only about adding GPU capacity. It also depends on data access patterns, governance, and how teams standardize the route from experiment to production.

Hiring signals put platform engineering at the center

CIO’s March coverage of Robert Half Technology’s 2026 IT salary report quantifies the talent-market pressure. In the report’s priorities list, AI and ML rank at 45% and IT operations and infrastructure at 36%. IT governance and compliance are listed at 25%, and cloud architecture and operations at 24%, according to CIO.

Compensation is often the first friction point for procurement and HR. In CIO’s summary of a Robert Half report, the median salary is $170,750 for an AI/ML engineer, compared with $98,000 for a systems administrator. The same CIO summary lists DevOps engineers at a $145,750 median and network/cloud engineers at a $132,000 median.

That gap pushes a pragmatic choice inside cloud programs: when expensive roles are hard to hire, standardization becomes part of the labor plan. Internal developer platforms, opinionated CI/CD templates, policy-as-code, and shared observability are not just architectural preferences. They cut down on one-off implementations that require $170K talent to operate and maintain.

Another Foundry data point, reported by CIO, aligns with that trend. Foundry found 36% of companies have added AI/machine learning engineers as part of their cloud investments. That suggests organizations are building in-house capability to deliver AI, rather than relying only on data-science hires and assuming the rest will follow.

Upskilling, adoption measurement, and the last mile of rollout

The Robert Half report also highlights a constraint that does not show up in architecture diagrams: readiness. According to CIO’s reporting of Robert Half’s findings, only 7% of leaders said they have the capabilities needed to complete prioritized projects, and 65% said they expect to upskill existing team members to close skills gaps.

That tension, high market pay for in-demand roles alongside an upskilling-heavy plan, often leads operators to look for ways to make learning repeatable in day-to-day work. BankInfoSecurity’s CIO.inc case study on Ferring Pharmaceuticals offers an application-layer example: Ferring used Icertis for contract lifecycle management and adopted Whatfix’s digital adoption platform to guide users through CLM workflows.

The point is not CLM itself. It is that some enterprises are treating adoption as a capability that needs dedicated tooling, because otherwise “upskill” can turn into a training calendar and a growing queue of support tickets.

Public-sector playbooks: staff governance and the operating model

Two government examples show how agencies are operationalizing cloud and data modernization through org design and cross-agency agreements.

Nextgov/FCW reported in April that the Pentagon named five senior leaders within the Office of the Department of Defense CIO as it works to oversee the department’s technology efforts. Nextgov said the announcement included roles such as a new chief of staff and a special advisor focused on organizational change and business process re-engineering.

Enterprise takeaway: in large environments, leaders are explicitly adding senior roles tied to organizational change and business process re-engineering around technology programs, a sign that portfolio governance, reuse, and execution across federated teams are central challenges.

Government Technology’s reporting on Illinois’ data-sharing agreement shows a smaller-scale version of that governance pattern. Illinois completed an enterprise memorandum of understanding across 13 state agencies to share data. In the same account, state CIO Hardik Bhatt said he brought 15 agency lawyers together for the effort, and he cited Indiana’s experience, where it took 18 months to get agencies to agree to share data. Bhatt’s takeaway was that doing one enterprise agreement moves faster than negotiating multiple one-off agreements, according to Government Technology.

For private-sector operators dealing with fragmented data ownership across business units, it is a useful analog. When the goal is interoperability, decision rights and contract structure can matter more than which ETL tool gets selected.

How this maps to 2026 operating plans for cloud, AI, and finance

• For CIO and infrastructure leaders: map your top 10 cloud workloads to the roles you actually need to run them. If delivery depends on a handful of AI/ML engineers priced at the Robert Half median, standardize the platform path so those specialists are not also doing glue work (per CIO citing Robert Half).
• For finance and procurement: treat the Foundry 26% cloud budget expectation as a forcing function to mature showback/chargeback and unit-cost reporting. If a cloud program cannot explain cost per workload, it will struggle to defend that budget share over multiple planning cycles (per CIO citing Foundry).
• For app owners rolling out new cloud and SaaS workflows: include digital adoption tooling in the rollout plan alongside training completion. Ferring used Whatfix’s digital adoption platform to guide users through Icertis contract lifecycle management workflows (per BankInfoSecurity).
• For data and governance leaders: inventory where interoperability is blocked by agreement structure. Illinois’ one-enterprise-MOU approach is a reminder that even fast technical architectures can stall if every domain negotiates separately (per Government Technology).