AI infrastructure spending set to nearly double to $42.3 billion in 2026: Gartner 

As per the report, total AI-optimized IaaS (Infrastructure as a Service) in 2025 was at $21.5 billion. In 2026, it increased to $42.2 billion. This is 96.4% increase from the previous year. Gartner expects the figure to rise further to $66.1 billion in 2027, marking a 56.5% increase from 2026.

IaaS market is also expected to grow. Global IaaS spending is projected to reach $287.3 billion in 2026, compared with $222.2 billion in 2025. Gartner expects spending to climb to $359.9 billion in 2027.

AI inference is becoming a bigger priority

AI inference is expected to account for a larger share of infrastructure spending as companies increasingly use AI in everyday business operations.

Gartner estimates that companies will spend $23.3 billion on inference workloads in 2026, compared with $19 billion on AI training. This means inference will make up around 55% of AI-optimised IaaS spending during the year. Its share is expected to increase to 59% in 2027.

The rise of agentic AI is one of the factors behind this growth. These systems can perform a series of tasks with limited human intervention, increasing the amount of computing power required.

“This growth is driven by continued demand for infrastructure to support large language model (LLM) training and the rapid operationalization of AI across enterprise applications and workflows,” said Hardeep Singh, senior Principal Research Analyst at Gartner.

Companies are also moving beyond developing AI models and using them directly in customer-facing products and internal operations.

“As organizations shift from model development to production-scale deployment, fine-tuned and domain-specific models (DSMs) are increasingly integrated into customer-facing and operational systems, requiring continuous, real-time execution rather than periodic training,” said Singh. “This shift is accelerating the cloud consumption patterns and creating sustained demand for AI-optimized infrastructure.”

AI could become part of almost all IT work

Gartner also expects AI to become closely involved in almost every area of IT by 2030.

The firm said that, by then, CIOs expect 75% of IT work to be done by people working with AI, while AI could independently handle 25% of IT work. Gartner’s forecast is based on a survey of more than 700 CIOs conducted in July 2025.

However, Gartner said simply having access to AI will not guarantee that companies benefit from it. Businesses will also need employees who are prepared to use the technology effectively.

“Gartner has been guiding CIOs and IT executives on their AI journeys for many years. In 2023, we showed them how to shape their AI ambition. Last year at IT Symposium/Xpo, we explained how to pace themselves in the AI outcomes race. This year, we’re mapping out the right path for them to take so they can go all-in on AI value,” said Gabriela Vogel, VP Analyst at Gartner.

“While not all AI is ready to deliver value, humans are even less ready to capture value,” said Rob O’Donohue, VP Analyst at Gartner. “AI readiness means AI can help you find value and effectively meet the needs of specific use cases. Human readiness is about whether you have the right workforce and organization to capture and sustain AI value.”

AI is expected to reshape jobs

Gartner expects the overall impact of AI on global employment to remain neutral through 2026. By 2028, the firm expects AI to create more jobs than it eliminates.

Companies, however, are likely to change the way they hire and use their existing workforce as AI becomes more capable.

“AI is not about job loss. It’s about workforce transformation. CIOs should start transforming their workforces by restraining new hiring (especially for roles involving low-complexity tasks) and by repositioning talent to new business areas that generate revenue,” said Vogel.

Some routine skills are also expected to become less important as AI takes over or assists with tasks such as summarising information, finding information and translation.

“AI will make some skills, such as summarization, information retrieval and translation, less important, as AI is ready to automate or augment these tasks,” said O’Donohue. “But AI also creates a need for entirely new skills. These AI skills are fundamentally different from most skills. Where skills were traditionally about doing tasks better, AI skills are about making you better — a better motivator, a better thinker and a better communicator.”

Gartner said companies will also need to ensure that employees do not become overly dependent on AI. Regular assessments could help ensure workers continue to retain important skills instead of allowing them to weaken through lack of use.

AI investments still come with challenges

Despite the rapid adoption of AI, many companies are yet to see clear financial returns from their investments.

In EMEA, 73% of CIOs said their organisations are either breaking even or losing money on AI investments, according to Gartner.

The firm said businesses need to look beyond the upfront cost of AI tools. Training employees, managing changes within the organisation and other related expenses can add significantly to the overall cost.

Gartner also said companies should distinguish between AI capabilities that are already mature and those that still need improvement. Search, content generation, code generation and summarisation are among the capabilities it considers ready for use.

AI accuracy and AI agents, on the other hand, still present challenges. Gartner recommended that companies focus on AI agents capable of making decisions and handling specialised tasks rather than relying only on conversational systems.

Choosing the right AI provider

The choice of AI provider will depend on what a company wants to achieve. Large organisations planning widespread AI deployments may benefit from hyperscalers because of their ability to provide infrastructure at scale.

For specialised industry applications, startups may offer more focused AI solutions and expertise. AI research and development companies, meanwhile, can provide access to newer technologies, although they may not have the infrastructure scale required by large enterprises.

Gartner also said companies should consider where their AI infrastructure and data are located and how much control they have over them. It described this as an important part of making AI-related technology decisions.

The firm has developed a framework to help organisations assess whether their technology, workforce and other resources are ready to support their AI plans. The aim is to help businesses move beyond adopting AI for its own sake and focus on areas where it can deliver measurable value.