Data 'Trust Gap' Threatens Autonomous AI, New Report Finds

PALO ALTO, CA – April 15, 2026 – A new era of artificial intelligence is dawning, one where AI systems can act independently to manage supply chains, execute financial trades, and resolve IT issues. But as enterprises race to adopt this “Agentic AI,” a fundamental crisis of trust in the underlying data threatens to derail progress before it truly begins, according to a new global study.

The AI Trust Gap Report, released today by data management leader Denodo, reveals that while businesses are eager to deploy autonomous AI agents, their existing data infrastructure is unprepared for the immense responsibility. The comprehensive study, conducted by Arlington Research, highlights a stark disconnect between the real-time, high-quality data that agentic systems require and what most organizations can currently provide, creating a dangerous “trust gap” that could stall innovation and expose firms to significant risk.

From Answering Questions to Taking Action

Unlike the now-familiar generative AI models that respond to prompts, Agentic AI represents a significant leap forward. These systems are designed to be proactive, capable of setting goals, creating multi-step plans, and executing complex tasks across various applications with minimal human oversight. Industry analysts have taken note, with Gartner recently placing AI agent development platforms at the “Peak of Inflated Expectations” on its Hype Cycle, predicting mainstream adoption within five years.

Use cases are already emerging across sectors. In finance, agents analyze market data to autonomously adjust investment portfolios. In HR, they automate complex employee onboarding and transition workflows. For IT departments, they promise proactive incident resolution, identifying and fixing problems before users are even aware of them. This shift from AI as an analyst to AI as an actor fundamentally changes the stakes. An incorrect answer from a chatbot is a nuisance; an incorrect action from an autonomous agent can have immediate and severe operational consequences.

The Cracks in the Data Foundation

The Denodo report quantifies the chasm between this autonomous future and today's data reality. The findings paint a picture of an enterprise data landscape that is fragmented, slow, and insecure—a treacherous foundation for systems designed to make independent decisions.

According to the research, the challenges are widespread and deeply technical:

• The Need for Speed: An overwhelming 66% of organizations state that for AI-driven insights to be trustworthy, the data must be accessed in real time. Stale or delayed data is simply not an option when an AI agent is about to trigger a purchase order or reconfigure a network.

• The Search for Context: Data without context is meaningless. 63% of respondents identified “finding relevant data” within a specific business context as a primary barrier to deploying AI. Agents need to understand not just the data points, but how they relate to a business process, a customer journey, or a compliance rule.

• The Security Paradox: As AI agents require access to numerous systems to perform tasks, they also expand the potential attack surface. Yet, 67% of organizations report struggling to maintain consistent security and access controls across their disparate data sources, a critical vulnerability for safe autonomous operations.

• The Scale and Complexity Barrier: The average enterprise AI initiative now pulls from over 400 different data sources, with 20% of organizations grappling with more than 1,000. This massive, siloed data estate makes creating a unified, reliable view of information a monumental challenge.

“AI is rapidly shifting from systems that merely answer questions to systems that take autonomous action, and this transition changes the data requirement entirely,” said Dominic Sartorio, vice president of Product Marketing at Denodo. “When an AI agent triggers a business outcome, there is zero room for stale or ungoverned data.”

A Crisis of Architecture, Not Algorithms

The report strongly concludes that this trust gap is not a failure of the AI models themselves, but a direct reflection of a broken underlying data architecture. The long-standing principle of “garbage in, garbage out” is amplified in the world of autonomous agents. As one industry analyst noted, “Organizations are trying to run a Formula 1 engine on unrefined crude oil. The failure isn't the engine; it's the fuel.”

This data crisis is the most critical of several hurdles facing agentic AI, including immense compute requirements, cultural resistance, and the operational complexity of monitoring non-deterministic systems. However, experts agree that without a trusted data foundation, none of the other challenges can be effectively addressed. Traditional data governance, which often relies on slow, manual, pre-deployment approvals, is wholly inadequate for the dynamic, continuous nature of agentic AI. These systems require a new paradigm of live, adaptive governance that can monitor and control data access in real time.

Building a Bridge to Trusted Automation

To close the trust gap, organizations are turning to a modern approach known as logical data management. Instead of undertaking costly and time-consuming projects to physically move and consolidate data into a single repository, a logical data fabric creates a unified, virtual layer that provides a real-time, governed view of data, no matter where it resides.

This approach directly tackles the core issues identified in the report. It enables real-time access to hundreds of disparate sources, enforces consistent security and governance from a central point, and uses AI-powered catalogs to help discover data in its proper business context. Companies across the data management landscape, including Ataccama and Informatica, are racing alongside Denodo to provide platforms that can deliver this AI-ready data foundation.

By building this logical bridge over their fragmented data estates, businesses can provide their AI agents with the reliable, timely, and context-rich information they need to act confidently and safely. For organizations looking to move beyond AI experimentation and unlock the transformative potential of automation, addressing the data trust gap is not just a technical priority—it is the strategic imperative of our time.