95 percent of companies are postponing AI projects due to infrastructure hurdles
A new global study by Cloudera shows that outdated data architectures significantly hinder the scaling of AI in enterprises. Nearly all of the organizations surveyed had to postpone or cancel AI initiatives due to governance, compliance, or regulatory issues. According to the study, the solution lies in hybrid data architectures.
Artificial intelligence is no longer just a pilot project in the corporate world—but in many places, the infrastructure isn’t keeping pace. This is shown by Cloudera’s new global study, «The Great AI Re-Architecture,» which was published on August 11, 2026. The survey polled 1,500 enterprise architects, cloud infrastructure managers, and data architects from companies with at least 1,000 employees across nine markets worldwide.
AI Is the Norm—But the Infrastructure Isn't
Although 77 percent of the companies surveyed are already actively using AI, nearly all (95 percent) had to postpone or cancel AI initiatives last year due to challenges related to data governance, compliance, or regulation. At the same time, 72 percent say their data architecture needs a fundamental overhaul to meet future AI requirements.
Three-quarters (75 percent) of respondents report that AI integrations have already changed their company’s data storage and architecture practices. 84 percent report increased infrastructure costs due to AI workloads. These figures make it clear that companies are rethinking not only where their data is stored, but also how it is managed, controlled, and made available to AI systems.
Governance Is Becoming Critical Infrastructure
As AI scales up within companies, governance is becoming increasingly important from a strategic perspective. 73 percent of respondents believe that AI has made data governance more complex. More than half (55 percent) have postponed or canceled more than six AI projects in the past twelve months due to governance, compliance, or regulatory hurdles.
The increasing decentralization of data storage further exacerbates this challenge: Nearly all respondents (97 percent) report moving data between different environments at least once a month. Consistent governance across cloud, private cloud, on-premises, and edge environments is therefore essential for the secure scaling of AI.
«The current AI era is forcing companies to rethink the fundamentals of their technology infrastructure,» says Sergio Gago, Chief Technology Officer at Cloudera. «Many organizations are realizing that architectures designed for traditional analytics are not built for the scale, governance, and flexibility that AI requires today. Success depends on creating a data foundation that allows companies to deploy AI where it makes the most sense, without compromising on control or security.»
Hybrid Architectures as the New Corporate Standard
To balance performance, governance, costs, and flexibility, companies are increasingly turning to hybrid architectures. Two-thirds (66 percent) of respondents migrated AI workloads from public cloud environments back to the private cloud or their own data centers last year. One-quarter (25 percent) plan to prioritize a hybrid-first architecture over the next two years.
Instead of relying on a single deployment model, companies are investing simultaneously in cloud, on-premises, edge, and hybrid environments. This brings AI optimization to the forefront—not just AI adoption. According to Cloudera, companies that modernize their data architectures and deploy AI where it makes the most sense are best positioned to deliver scalable and secure AI solutions with sustainable business value.
Source: cloudera.com
This article originally appeared on m-q.ch - https://www.m-q.ch/de/95-prozent-der-unternehmen-verschieben-ki-projekte-wegen-infrastrukturhuerden/
