Global IT spending reaches $6.31 trillion in 2026, a 13.5% jump from 2025. AI is the engine. Data centers, software, and security are all growing fast.
Gartner's April 2026 forecast put worldwide IT spending at $6.31 trillion for 2026, a 13.5% increase from the $5.43 trillion spent in 2025, which itself was a 7.9% increase from 2024. Those growth rates accelerating year over year is the part worth pausing on. IT spending growth is not just continuing at a steady pace. It is speeding up. And the driver is not hard to identify: the infrastructure buildout required to support generative AI is proving to be larger, more capital-intensive, and more sustained than most forecasters expected even two years ago.
To put $6.31 trillion in perspective: Germany's GDP in recent years has been roughly in the $4 to $4.5 trillion range. Global IT spending now exceeds that by a substantial margin. This is not a niche industry figure. It is one of the largest coordinated capital deployments in the global economy, and it is growing at rates that suggest the AI investment cycle is still in an early phase rather than approaching saturation.
Data center systems are one of the clearest indicators. Gartner's April 2026 forecast projects data center systems spending will exceed $788 billion in 2026. That category includes the servers, storage, networking equipment, and power and cooling infrastructure that AI workloads require. Training a large language model or running inference at scale is computationally intensive in ways that standard enterprise IT was not. The chips are different, the networking is different, the power density is different, and the cooling requirements have become an engineering problem in their own right. You cannot retrofit a traditional data center for dense GPU compute without significant physical infrastructure investment. The $788 billion figure reflects how seriously the market is taking that infrastructure requirement.
The compound effect here is worth thinking through carefully. AI infrastructure requires more servers. More servers require more power, more cooling, and more networking. More power requires utility infrastructure upgrades in regions where the grid is already stressed. More cooling at high power density requires new approaches to heat management that are more sophisticated than what data centers have traditionally used. Every layer of the stack is scaling simultaneously, and each layer creates demand for the next. This is why data center systems is one of the fastest-growing segments in IT spending right now, not because enterprises are buying more file servers, but because the compute requirements of AI workloads are physically and structurally different from the workloads that preceded them.
Software is growing too, and here Gartner's forecast is explicit: generative AI continues to drive outsized gains in software spending. This connects to something I think is underappreciated in coverage of AI investment. The spend on AI is not just happening in the infrastructure layer. It is showing up in software licensing, in AI-enabled features being added to platforms organizations already use, in new AI-native tools that are replacing older software categories, and in the development costs of building AI capabilities into existing products. The software layer of AI investment is distributed across the entire enterprise software market in a way that makes it harder to see than the data center buildout, but it is substantial.
The cloud infrastructure segment reflects this same dynamic. In its November 2024 forecast, Gartner projected IaaS end-user spending growing at 25.6% and PaaS at 20.6% in 2025, with cloud infrastructure and platform services reaching $301 billion. Cloud is the delivery mechanism for much of the AI compute that organizations are buying. Whether it is fine-tuning models, running inference APIs, or storing and processing the training data that AI systems require, cloud infrastructure is absorbing a significant share of AI-related investment alongside owned data center spending.
Information security spending tells a parallel story. Gartner forecasts worldwide end-user spending on information security at $213 billion in 2025. This is not coincidental with the AI surge. More AI workloads mean more attack surface. AI systems introduce new vulnerabilities, including prompt injection, model inversion, training data poisoning, and supply chain risks specific to AI components. Organizations deploying AI at scale are discovering that their existing security frameworks were not designed with these threat vectors in mind. The $213 billion security figure is partly a response to the same technology wave driving everything else in the IT spending forecast.
From an IS research perspective, the $6.31 trillion figure raises a question that the literature on IT investment and firm value has been wrestling with for decades: does this spending create commensurate value? The productivity paradox, Solow's observation that computers appear everywhere except the productivity statistics, has been revisited many times. My read is that we are at an early stage where the infrastructure is being built and the productivity gains are lagging, which is normal for transformative technology cycles. I wrote about why the productivity paradox still applies to modern IT investments in a post on Solow's paradox and what it means for AI. The $6.31 trillion question is whether this cycle will be different enough from previous ones that the gains show up faster and more broadly than past technology waves would predict. I am genuinely uncertain. The infrastructure buildout is real. The organizational capability to use AI effectively is lagging behind it. That gap is where the interesting research is.