Although the adoption of AI has become widespread, legacy data architectures increasingly limit organisations' ability to scale securely and efficiently. This is the main conclusion of the study The Great AI Re-Architecture, commissioned by the big data and data analytics company Cloudera and conducted by Wakefield Research. Thus, 84% of Spanish organisations believe that their data architecture needs a thorough review, a figure that is above the global average (72%).
After surveying 1,500 technicians and IT managers globally, the report reveals that AI is causing a reconfiguration of enterprise infrastructure. As organisations integrate AI into their business, it becomes essential to have an infrastructure designed to scale these solutions. Consequently, 6 out of 10 (63%) say that the implementation of artificial intelligence has changed their data storage and architecture practices to a greater or lesser extent.
Companies are rethinking not only where data resides but also how it is managed, governed, and made available to AI systems. In this sense, the massive transition from legacy data architectures to hybrid environments is becoming crucial.
Nearly half (49%) of Spanish organisations have moved AI workloads from the public cloud to private cloud or on-premise infrastructure over the past year. And, looking ahead to the next two years, these organisations intend to combine various infrastructure approaches, with a predominance of public cloud and hybrid models: 35% plan to increase their cloud spending, and 31% will prioritise a hybrid approach, while 17% will invest more in edge and 16% in on-premise.
Governance and compliance, the main hindrance for AI projects
Nearly half (49%) of Spanish companies believe that maintaining effective governance has become more difficult with the integration of AI, although this figure is below the 65% in the EMEA region. In fact, security, governance, and compliance are cited as the main reasons for changes in AI infrastructure by 32% of respondents.
In this context, marked by stringent European regulations, 98% of Spanish organisations have delayed or cancelled AI projects in the past twelve months due to governance and regulatory compliance challenges.
“Many companies are discovering that architectures designed for traditional analytics were not built for the governance and flexibility that AI demands today. Success will depend on building a solid data foundation that gives organisations the freedom to run AI where it makes the most sense, without sacrificing control or security,” explains Sergio Gago, CTO of Cloudera.