Cloudera survey finds AI projects hit governance strain

Cloudera has published a global survey showing that 95% of enterprises have delayed or cancelled AI projects because of infrastructure and governance constraints. The findings are based on responses from 1,500 technology specialists across nine markets.

The results point to growing pressure on enterprise systems as businesses expand artificial intelligence beyond pilot programmes. Among respondents, 77% are actively using AI, while 72% believe their current data architecture needs a significant overhaul to meet future AI requirements.

The research drew on responses from enterprise architects, cloud infrastructure leads and data architects at organisations with at least 1,000 employees in most markets. It covered the US, Canada, Brazil, South Africa, Spain, the UK, Singapore, India and Japan.

Infrastructure strain

The survey found that 75% of respondents said AI integrations had changed their organisation's data storage and architecture practices. Another 84% reported higher infrastructure costs linked to AI workloads.

That combination is pushing companies to reconsider not only where they store data, but also how they manage and govern it as AI systems draw on information across multiple environments. Legacy architectures built for conventional analytics are proving difficult to adapt to the demands of broader AI use, Cloudera argued.

"This current era of AI is forcing organisations to rethink the foundations of their technology infrastructure," said Sergio Gago, Chief Technology Officer, Cloudera. "Many enterprises are discovering that the architectures built for traditional analytics weren't designed for the scale, governance, and flexibility AI demands today. Success will depend on building a data foundation that gives organisations the freedom to run AI wherever it makes the most sense, without compromising control or security."

Governance issues

Governance emerged as one of the main reasons projects are being held back. Nearly three-quarters of respondents, or 73%, said AI had made data governance more complex.

More than half, or 55%, said they had delayed or cancelled more than six AI projects over the past year because of governance, compliance or regulatory issues. The survey also found that 97% move data between environments at least once a month, adding to the difficulty of applying consistent rules across cloud, private cloud, on-premises and edge systems.

Those patterns suggest governance is becoming a central operational issue rather than a supporting function. As data moves more frequently between systems, companies face a greater challenge in maintaining oversight and meeting internal and external requirements while still deploying AI tools at pace.

Hybrid shift

The survey also found a marked shift towards hybrid technology models. Two-thirds of respondents, or 66%, said they had moved AI workloads from public cloud environments back to private cloud or on-premises infrastructure over the past year.

This reflects a broader effort to balance cost, control and operational requirements rather than rely on a single deployment model. One-quarter of respondents said they plan to prioritise a hybrid-first architecture over the next two years.

For companies managing sensitive or regulated data, the shift may also reflect concerns about data sovereignty and where AI processing takes place. The findings suggest businesses increasingly want the flexibility to place workloads in different environments depending on cost, governance or operational needs.

Cloudera also included an Australia and New Zealand perspective alongside the global findings, saying businesses in the region are dealing with the same mix of infrastructure, compliance and security demands seen elsewhere in the survey.

"Across Australia and New Zealand, leaders are navigating a complex web of AI priorities, including flexibility, data sovereignty, security, and compliance. The challenge is to achieve these goals without inhibiting innovation, which is putting a strain on existing infrastructure that was not designed to support AI at scale. To address this issue, organisations must rethink their data architectures, governance frameworks, and operating models. By establishing a trusted data foundation, businesses can harness the power of AI wherever it adds the most value, while maintaining control over their most valuable asset, their data," said Keir Garrett, Managing Director, Cloudera Australia & New Zealand.

The data adds to a broader picture of companies encountering practical barriers after initial AI roll-outs. While adoption rates remain high, the survey suggests that scaling those projects across large organisations is exposing weaknesses in older systems, especially where data is spread across multiple platforms and subject to different compliance demands.

Cloudera's sample focused on large organisations, with a lower minimum threshold in a small number of markets. The results indicate that, for many businesses, the challenge is no longer whether to use AI, but whether existing infrastructure and governance models can support it.