90% of companies expect a GenAI-infused cost increase

90% of companies expect a GenAI-infused cost increase

Nutanix’s Enterprise Cloud Index survey shows that organizations are massively betting on generative AI despite expected cost increases. As many as 90 percent of companies surveyed expect an increase in IT costs due to GenAI deployments. Meanwhile, 70 percent are counting on a return on investment within three years.

The survey, conducted among 1,500 IT and DevOps decision makers worldwide, shows that more than 80 percent of organizations have already developed a GenAI strategy. Exactly what this means will vary widely. In fact, other surveys show that many organizations are not yet past an experimental phase with GenAI.

Reading tip: Gartner: increase in IT spending in 2025 thanks to GenAI

Containerization as the foundation

One notable development is that nearly 90 percent of organizations are using containers for applications. This approach is seen as essential for the successful implementation of GenAI workloads. Nutanix is capitalizing on this with solutions that allow AI workloads to run both on-premises and in the public cloud. This responds to the 98 percent complaint among respondents that GenAI is difficult to implement with legacy systems.

“Many companies are at a tipping point with GenAI,” said Lee Caswell, SVP Product and Solutions Marketing at Nutanix. “Our ECI shows trends we also hear from customers, such as challenges related to scaling GenAI workloads from development to production, new requirements GenAI creates for data governance, privacy and visibility, and integration with existing infrastructures. A successful return on investment (ROI) from GenAI projects requires a holistic approach to application and IT infrastructure modernization, including containerization.”

Security remains top priority

For 95 percent of organizations, GenAI is changing priorities, with security and privacy at the top of the list. This is also reflected in specific concerns: 95 percent believe more needs to be done to secure LLMs. By the way, this can be interpreted in two ways. The LLMs themselves, after fine-tuning and domain-specific training, are themselves highly valuable as they regularly contain sensitive information, while their use can also lead to problems. After all, we have yet to rid ourselves of AI hallucinations, and even the most advanced models prove susceptible to malicious use.

Also read: DeepSeek is unsafe, and it’s got nothing to do with China

Investing in people and technology

The survey also highlights the importance of human capital, with 52 percent of respondents saying they need to invest in IT training, while 48 percent see recruiting new talent as necessary. Despite strong competition for GenAI expertise, 53 percent expect existing employees to develop AI skills organically.

Tip: Nutanix promotes multi-cloud collaboration in Europe via community platform