Despite growing interest and enthusiasm for Generative AI (GenAI), significant challenges are emerging that threaten the success of GenAI projects, according to a co-sponsored research report from Enterprise Strategy Group (ESG) and Hitachi Vantara, the data storage, infrastructure, and hybrid cloud management subsidiary of Hitachi, Ltd. (TSE: 6501). Surveying 800 IT and business leaders across the United States, Canada, and Western Europe, the report explores the critical role of data infrastructure for enterprise GenAI and the associated decisions underpinning successful implementation, finding that 97% of organizations with GenAI in flight view it as a top-five priority, with U.S. companies 35% more likely to say it was the top priority compared to European respondents.
Additionally, nearly two-thirds (63%) say that they have already identified at least one use case for GenAI. Despite the increasing pursuit of GenAI implementation, however, several factors pose serious risks for businesses:
Less than half (44%) of organizations have well-defined and comprehensive policies regarding GenAI.
Only slightly more than one-third (37%) believe their infrastructure and data ecosystem is well-prepared for implementing GenAI solutions; however, C-level executives were 1.3 times more likely to indicate that their infrastructure and data ecosystem is highly prepared, highlighting a notable disconnect.
61% of respondents agreed most users don’t know how to capitalize on GenAI, with 51% reporting a lack of skilled employees with GenAI knowledge.
40% of respondents agreed they are not well-informed regarding planning and execution of GenAI projects.
“Enterprises are clearly jumping on the GenAI bandwagon, which is not surprising, but it’s also clear that the foundation for successful GenAI is not yet fully built to fit the purpose and its full potential cannot be realized,” said Ayman Abouelwafa, chief technology officer at Hitachi Vantara. “Unlocking the true power of GenAI, however, requires a strong foundation with a robust and secure infrastructure that can handle the demands of this powerful technology.”
Building the Foundation for Enterprise GenAI
Data shows that organizations are actively seeking out lower-cost infrastructure options, but privacy and latency are also top factors in consideration. 71% of respondents agreed that their infrastructure needed to be modernized before pursuing GenAI projects - an overwhelming 96% of survey respondents prefer non-proprietary models, 86% will leverage Retrieval-Augmented Generation (RAG) and 78% cite some mix of on-premises and public cloud for building and using GenAI solutions. Over the long term, however, organizations expect the use of proprietary models to increase – six-fold according to the survey – as businesses gain expertise and seek to achieve competitive differentiation.
“The need for improved accuracy shows organizations prioritizing the most relevant and recent data gets incorporated into a Large Language Model, followed by the desire to keep pace with technology, regulations and shifting data patterns,” said Mike Leone, principal analyst at Enterprise Strategy Group. “Managing data with the right infrastructure will not only enable greater levels of accuracy, but also improve reliability as data and business conditions evolve.”
Drivers and Barriers to Adoption
The report found that several areas are driving companies to GenAI, as well as giving them pause. In terms of what’s driving enterprise investment in GenAI, the most cited use cases centered around process automation and optimization (37%), predictive analytics (36%), and fraud detection (35%). It’s therefore no surprise that improving operational efficiency was the area most cited for where businesses are seeing results; however, less than half (43%) have realized benefits up to this point.
When it comes to some of the top concerns and challenges being faced, more than four in five (81%) of respondents agreed on concern around ensuring data privacy and compliance when building and using applications that leverage GenAI, while 77% agreed that data quality issues needed to be addressed before accepting the results of GenAI outputs.
