5 ways enterprise AI will transform IT infrastructure in 2025

Unexpected transition in IT was the theme of 2024. Adjustments to VMware's licensing model surprised many organizations. New, sometimes undefined, artificial intelligence projects appeared at the top of IT priority lists. The resulting effects have led IT organizations worldwide to become even more hyperfocused on cost. The result? 2025 will be the year your organization chases the promise of enterprise AI and invests to shore up data resilience and cyber recovery capabilities, all while you learn how far you can stretch your IT budget dollar.

With these defining themes for 2025 in mind, let's discuss what the next 12 months will bring.

1. 2025 will be the year to get artificial intelligence in production -- chatbots don't count -- or risk being left behind. Over the next 12 months, I expect early movers in AI will emerge across every industry. At that point, the competitive differentiation will begin, and the laggards will be forced to play catch-up. A good practice when evaluating initial AI use cases is to start with something internal, small and able to provide a quick win to prove the value of AI to executive leadership. This project will also help to identify blind spots or best practices along the way.

To that end, internal or externally facing agents and/or chatbots can serve as a strong early use case. But the incremental business value of these tools is often difficult to measure, and any competitive advantage will quickly evaporate. Businesses must look beyond chatbots to create differentiated business value with AI. Multiple technology partners and providers, such as Nvidia, Dell Technologies, IBM and the major public cloud providers, can advise organizations on use cases as well as technology. While it is still relatively early in the age of AI, 2025 is the year production-level AI will begin to matter competitively and financially. I would not want to enter 2026 without AI in production. While most players in your industry might not have AI implementations ready by then, there is a strong likelihood someone will.

I would not want to enter 2026 without AI in production.

2. By the end of 2025, AI observability becomes a top 10 IT priority, fueling increased storage investment. If we assume AI-powered apps will be high-priority competitive differentiators, then minimizing risk, ensuring an optimal experience, addressing drift, reducing hallucinations and minimizing bias or data exposure for them will all become top priorities. The result will be increased investment in AI observability technologies designed to monitor and track the elements that best define the optimal AI experience. As a result, the capacity required to collect all the data necessary to observe those AI-powered apps should scale quickly. With application observability in general, 69% of organizations agreed that their observability data is growing at a concerning rate, according to a survey from Informa TechTarget's Enterprise Strategy Group. It is only logical to expect AI observability environments to grow in a similar fashion.

3. Cost pressures will accelerate on-premises modernization efforts. The need to scale infrastructure to support AI will span both on- and off-premises investments. While early AI deployments have favored public cloud, on-premises environments are seeing growing interest as an option to reduce the cost of infrastructure for production AI. According to research from Enterprise Strategy Group, 78% of organizations agreed that they would prefer to run their AI applications on-premises. The result is that some portion of AI investment will focus on data center modernization.

In addition, organizations are still defining how best to respond to VMware's revised licensing model. While some organizations are expected to embrace VMware Cloud Foundation, others are investigating other hypervisor options. This has led to increased interest in options from HPE, Nutanix, Verge.io and Red Hat.

The combined effects of both AI and shifts in the virtualization space will place more price pressure on near-term budgets. This price pressure will likely fuel an uptick in adoption of consumption-based models for on-premises infrastructure or on-premises infrastructure as a service, such as the options from Dell, HPE, NetApp, Hitachi Vantara, Pure Storage or Lenovo. These as-a-service pricing models can help organizations defer the cost of growth into future quarters while also reducing the burden on internal personnel.

4. You will be targeted by ransomware, reaffirming the importance of optimizing your strategy for data resilience and cyber recovery. History repeats itself, and disruptive technologies are often first adopted by groups that operate outside the law. Generative AI is poised to make cyberattacks more frequent and more formidable. If you have not invested in adequate tools to identify attacks, respond to them and recover from them, the clock is ticking. The lines between data protection, data security, data resilience and cyber recovery continue to blur as vendors work to respond to increased threats. It isn't just about recovering data, it's about how you can deliver resilient business operations.

The recent acquisitions of Clumio and Appranix by Commvault, along with the integration of Cohesity and Veritas, highlight just how important innovation will be in this space. In addition, recent announcements from players such as Veeam, Dell, HPE and Rubrik serve as additional examples of how rapidly vendors in this space are working to address the evolving threat landscape.

5. Every solution will claim to have AI inside, and the responsibility will fall on you to figure out which AI is better. Every technology vendor is working to integrate AI capabilities into its portfolio. Those that fall behind, however, will likely be forced to AI wash their marketing efforts to provide the perception of keeping pace. What does this mean to you? It means that when a vendor tells you it has integrated AI features, you must apply added scrutiny. How was it trained? What data was used? How often is it updated? What is the experience?

This breaks down how I expect the AI dominos will fall over the next 12 months. In the upcoming AI era, the investments and innovations in 2025 might not decide the winners, but for those that stand still, it might define the losers.

Scott Sinclair is Practice Director with Informa TechTarget's Enterprise Strategy Group, covering the storage industry.

Enterprise Strategy Group is a division of Informa TechTarget. Its analysts have business relationships with technology vendors.