The challenge

As AI moves from experimentation into operational decision-making, organisations are facing a new problem: trust.

While Agentic AI systems are increasingly expected to act autonomously, interact with live systems, and support real-time decisions, many organisations lack the data foundations needed to make those systems reliable, governed, and scalable.

Denodo wanted to understand where businesses are struggling most, and what’s preventing trustworthy AI from moving beyond pilot stages into production environments.

The approach

Arlington Research partnered with Denodo to deliver The AI Trust Gap Report, based on a global survey of 850 enterprise leaders responsible for AI initiatives across multiple industries and regions.

The research explored the operational realities of enterprise AI, focusing on:

• Real-time data access
• Data quality and relevance
• Governance and guardrails
• Performance and scalability
• Distributed enterprise data environments

The report examined how organisations are approaching AI readiness, and where gaps between ambition and execution are emerging.

The insight

The findings reveal that many organisations are attempting to scale AI without the operational data foundations required to support trustworthy outcomes.

The research identifies several major gaps:

• A live data gap, where AI systems require real-time situational awareness, but existing architectures are designed for historical analytics
• A relevance gap, where organisations struggle to identify trustworthy, contextually relevant data
• A guardrail gap, where governance, access controls, and compliance remain difficult to enforce consistently across distributed systems

The report also highlights the growing complexity of enterprise AI environments. Many organisations are drawing on hundreds of separate data sources, increasing the challenge of maintaining consistency, governance, and performance at scale.

The impact

The report gives enterprise leaders a clearer understanding of why AI initiatives often stall between experimentation and production.

It highlights the growing importance of live operational data, semantic consistency, governance, and scalable architectures in enabling trustworthy AI systems.

For Denodo, the research reinforces its position as a leader in logical data management and enterprise AI readiness. The report also achieved strong industry visibility, generating 200+ LinkedIn mentions and sparking discussion among enterprise AI and data leaders.

Download the report