The appetite for AI in the market has never been greater. According to Gartner, over 90 percent of CIOs globally are increasing funding in AI, making it the fastest‑growing area of enterprise technology spend. As organizations look to integrate AI-powered workflows, from real-time analytics to personalized customer experiences, this ambition is accelerating investment in data initiatives.
Additional research shows that enterprises now spend an average of $29.3 million per year on data programs – which encompasses data movement, ingestion and preparation tooling, recurring cloud ingest and compute costs, and the internal engineering capacity required to keep pipelines running.
While this shift in spend mirrors the demands of scaling AI (organizations with successful AI initiatives invest up to four times more in data and analytics foundations), higher budgets do not automatically result in high‑quality data. Many businesses continue to miss out on the transformative impact of AI, held back by underlying weaknesses in their data architecture that slow delivery and limit returns.
Almost two thirds of data initiatives are underperforming
Despite unprecedented levels of investment, the majority of enterprise data initiatives continue to underperform – with 73 percent of organizations reporting their data initiatives are falling short of expectations. At the same time, nearly 62 percent report low levels of data maturity, pointing to a persistent gap between what organizations want their data and AI initiatives to deliver, and what their infrastructure is equipped to support.
Weak data foundations constrain innovation and carry measurable consequences for enterprise performance. In large organizations, downtime caused by data pipeline failures now exceeds 60 hours a month, disrupting productivity and costing an estimated £50,000 per hour in business impact. Data teams are also affected, as they spend over half of their engineering capacity on pipeline maintenance, rather than advancing new use <a href="https://myappsplus.com/best-iphone-18-pro-cases-now-available-for-purchase9to5mac/” title=”Best iPhone 18 Pro cases now available for purchase9to5Mac”>cases.
Open Data Infrastructure as the foundation for AI
Beyond the day‑to‑day costs of downtime and maintenance, the deeper impact of unreliable data foundations is consistent disruption of AI initiatives. For AI systems to thrive, organizations need democratized, interoperable data programs, where access to data is fast, governed and reliable. In response, Open Data Infrastructure (ODI) has emerged as the foundation for AI.
ODI is an architectural approach that gives organizations greater control over how data is accessed, moved and used, by allowing tools and platforms to work together through shared, open standards. Instead of relying on tightly coupled, proprietary systems, ODI is built on a modular, standards‑based foundation that separates storage from compute, enabling each layer to evolve independently.
As data and AI workloads continue to grow, this creates a unified data environment where analytics and AI can scale more efficiently.
ODI is also emerging as a direct challenge to vendor lock‑in. The industry is seeing a shift towards data becoming more restricted, both technically and commercially. Often, these constraints show up as hidden costs or dependencies that push companies toward specific walled-garden ecosystems.
This problem is amplified when AI entities become an organization’s primary data users. Indeed, studies suggest that non-human entities are present in modern enterprises at a ratio of 82:1 compared to humans.
For AI agents to work effectively alongside human users, a sharedchine learning models and AI agents may all draw from the same underlying data, but often operate in separate environments with different definitions and models
When those definitions drift, the result can be misaligned decisions, unreliable AI outputs and additional engineering overhead. ODI helps address this by giving every system, human or automated, a consistent view of the business.
Furthermore, AI agents generate exponentially more queries than humans, but closed ecosystems often route them through the same expensive compute infrastructure. Agents can only optimize for cost – opting for cheaper compute engines when appropriate – when open architectures afford them the opportunity to choose. And the cost considerations don’t stop there.
Organizations using legacy systems pay significantly more per data pipeline, which, when multiplied by the hundreds of pipelines at enterprise scale, adds up to a significant, ongoing expense.
Modern data management: flexible, portable, trusted
As investment in AI tools continues to ramp up, organizations must think ahead to alleviate the strain on both budgets and engineering resources. They should ensure AI systems have consistent access to fresh, trustworthy and context-rich data while maintaining control of their data and architecture to avoid lock-in.
Those that prioritize open foundations will create the right conditions for innovation and enable their data teams to focus on delivering real business value, from predictive modelling and real-time analytics to faster agent production.
The impact is ultimately reflected in performance outcomes. Research shows that organizations with modern, managed and open data foundations are nearly twice as likely to exceed their ROI targets than those relying on legacy systems – evidencing the direct correlation between data maturity and measurable success of AI initiatives.
We’ve featured the best AI website builder.
This article was produced as part ofTechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.
The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here:https://www.techradar.com/pro/perspectives-how-to-submit
