Survey Suggests Corporate Leaders Are Underestimating The AI Infrastructure Challenges Ahead

Enterprise data center tenants are all-in on artificial intelligence, but a new survey indicated that top executives are overestimating their infrastructure’s AI readiness. 

Large corporations across nearly every industry are beginning to execute their AI roadmaps — formal gameplans to incorporate artificial intelligence into their products, services and internal processes.

Boards and chief executives are often the driving force behind this AI investment, according to a report published this week by colocation data center provider Flexential. Yet the analysis also suggests that these leaders are often unaware of the significant infrastructure hurdles and subsequent costs that stand in the way of executing on these AI roadmaps. 

"The stakes are high for AI infrastructure investments," Flexential CEO Chris Downie said in a statement. "Our survey shows that enterprise leaders are ready to execute on ambitious AI plans, with over half of leaders stating that C-level leadership is behind the pressure to rapidly adopt AI. Yet, many enterprises feel held back by significant challenges in their IT infrastructure, workforce, or organization's leadership."

Flexential surveyed 350 IT leaders at organizations with over $100M in annual revenue across a range of industries, from software and manufacturing to retail, healthcare and utilities.

Those surveyed reported significant barriers standing in the way of achieving their companies’ AI goals, with the most significant hurdles being insufficient infrastructure and staffing. 

82% indicated they encountered performance issues with their AI infrastructure in the last 12 months, identifying the top causes of these issues as bandwidth shortages, unreliable connections and difficulty scaling data center space and power. Organizations are also struggling to hire employees with the skills and expertise needed to support their AI infrastructure, with 91% reporting some kind of staffing shortage or skills gap.

Yet top-level executives at these companies may not be fully aware of these problems happening on the ground, according to the report’s authors. 

One-third of C-Suite respondents to Flexential’s survey believed their organizations hadn't encountered any of these performance issues with their AI infrastructure over the last 12 months, compared to 19% of directors and 8% of vice presidents who expressed similar optimism. Similarly, the percentage of C-suite respondents who said they had experienced no staffing or skills gap issues was more than triple that of more junior respondents.

This gap between how top executives and lower-level “boots on the ground” IT leaders view the state of their companies’ AI infrastructure may be a looming problem within firms in the early stages of their AI game plans, the report’s authors suggest. C-suite leaders may not have a realistic picture of their organizations’ infrastructure shortcomings, which will likely require significant additional investment to address.  

Enterprise spending on data centers and other digital infrastructure to support AI is on the rise.

While cloud and social media giants like Amazon, Microsoft and Meta account for most of a wave of AI infrastructure spending that is expected to pass $1T within five years, there is a surge in AI investment from enterprises looking to incorporate AI into their products and services or use it to improve internal processes. Companies are starting to spend more on infrastructure like colocation data center capacity and AI-focused private cloud providers.

The rising tide of demand is also apparent in Flexential’s report. Nearly all respondents indicated their company had a formal, documented AI roadmap, and close to 60% indicated that their AI roadmap included increasing IT infrastructure investment. 

“It's still early days, but enterprises are trying to ensure that they’ve lined up the resources to enable AI,” Downie told Bisnow. “That’s going to take some time. Companies are trying to figure out how they can enable it as fast as they can, but enterprises always move slower than technology.” 

The pressure driving these AI initiatives is often coming from the top down. According to Flexential’s analysis, a majority of respondents indicated that their chief executive or board was a primary force behind their organization's decision to adopt or develop AI applications. 

Across all levels of these organizations, IT leaders are almost universally supportive of these AI goals. There is widespread belief among those surveyed by Flexential that these AI roadmaps are necessary and crucial for the performance of the organization — not frivolous or hype-driven spending. 93% of respondents said there would be consequences for their company not achieving the goals in their AI roadmaps, from having to put new products on hold to losing market share to competitors.  

But even though IT leaders tend to be believers in their companies’ AI efforts, they have big concerns about the ability to execute them. One-third of respondents suggested their organizations are playing catch-up when it comes to building AI capabilities, while 46% expressed some level of doubt about their ability to execute on their AI roadmap.

At the same time, IT professionals told Flexential that their company’s AI infrastructure spending has come with greater scrutiny of cost, with more than 90% of respondents reporting growing pressure from leadership to minimize time-to-revenue for AI infrastructure. The report’s authors say this desire for quick returns could be on a collision course with infrastructure needs that are bigger or more costly than top C-suite leaders and corporate boards expect. 

“IT leaders have a board-level mandate to invest significant resources in executing their AI roadmaps. But while they’re enthusiastic about these efforts, IT leaders are less confident about achieving them,” the report’s authors wrote. “Pressure to minimize time-to-revenue clashes with the reality of the significant infrastructure investments required to support complex AI use cases moving forward — a key element of most organizations’ AI roadmaps.”