How businesses are cutting cloud spend to fund AI investment

According to the 2026 Flexera State of the Cloud Report, public cloud spend is estimated to be 29% wasted. However, 81% of respondents are already utilizing generative AI. Technology leaders are now facing the challenge of recovering wasted spend without impeding AI investment, data, and modernisation. Cloud was anticipated to reduce infrastructure costs and enhance agility; yet, many businesses are experiencing rising monthly bills, limited visibility into cost drivers, and mounting pressure to justify technology investments.

DVT, a contributor to the FinOps Foundation, collaborates with organisations to boost visibility, accountability, and efficiency across cloud environments. They utilise cloud architecture and engineering expertise alongside FinOps practices to identify immediate savings and optimise long-term cloud platform efficiency. Naicker points out that businesses typically manage cloud environments as traditional data centres, moving applications without redesign.

Resources remain sized for peak demand, workloads run inappropriately, and organisations often overlook cloud efficiencies.

South African organisations face additional challenges due to foreign currency exposure, ageing applications, and legacy platforms that consume resources without yielding proportional business value. Establishing where cloud spend is directed, who owns it, and which resources generate value is the initial step. Reviews of cloud environments often reveal over-provisioned resources, under-utilised infrastructure, unused licences, and development or test environments with unclear ownership.

The process typically focuses on discovering the first 20% of savings, providing immediate financial benefits while also revealing broader opportunities for application modernisation, process automation, and improved cloud efficiency. Cloud cost control does not solely belong to the finance department; engineering and architectural decisions significantly determine spend, and finance must understand if that spend supports organisational priorities.

Finance teams discuss in rands and dollars, whereas engineering teams speak in compute, storage, and clusters. The FinOps Framework offers a shared language helping both parties comprehend trade-offs and make better decisions. DVT facilitates communication by connecting technical usage and ownership to financial information and business priorities, enabling decisions on resizing, retiring, automating, or redesigning resources.

This approach enhances accountability and provides a clearer basis for deciding where cloud investment should increase, decrease, and how its return should be measured.

The cost challenge extends to AI. Flexera's 2026 State of ITAM Report indicates that 59% of organisations have experienced increased AI software spend, surpassing other cloud software, SaaS, and infrastructure as a service. Reducing unnecessary cloud costs allows organisations to allocate funds to AI, advanced analytics, and modern data platforms without solely increasing the technology budget.

The objective is not indiscriminate cost-cutting but reallocating funds from low-value resources to initiatives that can boost growth, service, or productivity. Legacy systems often support critical business processes, but organisations must also maintain the flexibility to invest in emerging technologies that create new growth opportunities.

As AI adoption accelerates, managing AI costs, such as GPU infrastructure, model usage, and token consumption, will become as crucial as managing traditional cloud infrastructure.