For data and analytics leaders navigating today’s fast-evolving technological landscape, artificial intelligence has fundamentally raised the operational bar for everything underneath it. Data quality, data integration, and above all, data governance are no longer viewed merely as compliance checkboxes or administrative chores. Instead, they have transformed into a genuine strategic foundation for modern enterprises. Organizations that have failed to treat data governance as a core strategic pillar are now beginning to feel that pressure acutely as they attempt to deploy AI at scale.
That sobering reality served as the central premise of a recent session at the Data Integrity & AI Forum, titled AI Is Stress Testing Data Governance: Are You Seeing the Same? The panel discussion brought together three industry experts who offered distinct yet complementary perspectives on the modern data challenge: Nicola Askham, widely known as the Data Governance Coach; Kevin Guyot, Vice President of Data Analytics at Lake Michigan Credit Union (LMCU); and David Woods, Senior Vice President of Global Services at Precisely.
The conversation highlighted several critical themes that data and analytics leaders must grapple with as autonomous systems and advanced machine learning models demand unprecedented levels of data integrity.
Why Data Governance Produces Trust, and What Happens Without It
One of the most striking signals emerging from the 2026 State of Data Integrity and AI Readiness report—published by Precisely in partnership with Drexel University’s LeBow College of Business, based on a survey of more than 500 data and analytics leaders—is the direct statistical relationship between formal data governance programs and organizational trust in data.
According to the research, data leaders working within organizations that maintain formal governance programs report high or very high trust in their data at a rate of 71%. In stark contrast, only 50% of organizations without a formal governance program in place share that same level of confidence. Without a structured governance framework, an organization’s trust in its own data is essentially a coin toss.
That lack of trust carries immediate and severe consequences once artificial intelligence enters the corporate ecosystem. Modern machine learning models and autonomous agents consume immense volumes of information, and when they ingest untrustworthy data, they produce flawed outputs at unprecedented scale and speed.
Kevin Guyot’s experience at Lake Michigan Credit Union cuts straight to the heart of why that trust gap exists across different institutions. For Guyot, the dividing line between organizations that genuinely trust their data and those that struggle with doubt is not merely a matter of having extensive documentation. Rather, it is determined by whether governance is actively embedded into how business decisions are actually made on a daily basis. When governance possesses true ownership and permeates day-to-day operations, organizational trust in data naturally follows.
David Woods of Precisely observed that the primary driver for corporate data governance has shifted significantly over the past two to three years. What was once considered an industry-specific concern concentrated primarily in heavily regulated environments such as financial services and healthcare has rapidly evolved into a universal corporate requirement. Woods noted that as organizations feed models bad or uncontextualized data, the resulting outputs erode trust at an alarming rate. Consequently, governance has transitioned from a routine compliance exercise into an absolute business necessity.
This urgency is further compounded by a notable confidence-reality gap revealed in the survey data. While an overwhelming 87% of data leaders describe their respective organizations as being fully AI-ready, 43% simultaneously cite data readiness as a top barrier holding back their AI success. Industry experts point out that this discrepancy is not a mere reporting inconsistency, but rather a fundamental measurement problem. Many enterprises are currently evaluating their technological readiness against outdated criteria. As organizations shift toward agentic AI—where autonomous systems make critical business decisions without direct human oversight—these underlying data readiness gaps quickly become acutely consequential.
Does Your Governance Program Have a Data Strategy Behind It?
While the survey indicates that 83% of organizations have some form of data governance program in place, the persistent citation of data readiness as an AI barrier suggests that the critical question is no longer whether governance exists, but whether that governance is actively performing the right work.
Live audience polling during the forum highlighted a consistent organizational hurdle: a widespread lack of clear ownership and accountability. David Woods pointed to a recurring pattern he encounters when clients seek external assistance, noting that data governance initiatives are frequently built entirely disconnected from a cohesive overarching data strategy. Organizations that treat governance and strategy as completely separate workstreams consistently struggle to articulate business value or point to outcomes where governance served as an accelerator for success. Conversely, the moment those two functions connect, measurable business results and heightened trust scores follow closely behind.
Industry expert Nicola Askham emphasized that it ultimately matters little how sophisticated a governance program looks on paper if no one within the broader business can clearly explain the specific problems it is designed to solve. During a recent coaching session, Askham recalled a client listing a "better understanding of our data" as a primary benefit of their governance initiative. Her blunt response to the client was simple: "So what?" Askham explained that while a deeper understanding of data will occur, it fails to address what business stakeholders actually care about. Aligning governance directly with business outcomes—asking precisely what business problems will be solved and what tangible value will be delivered—remains paramount.
At Lake Michigan Credit Union, Kevin Guyot illustrates what top-down alignment looks like in daily practice. Rather than beginning with abstract governance protocols, LMCU starts by identifying core business goals, priorities, and objectives, using governance merely as the mechanism that makes those desired outcomes achievable. By communicating the underlying "why" to operational teams—explaining how clean data allows for faster, more confident decision-making and reduces the hours spent reconciling conflicting reports—organizations can bridge the gap between technical data management and executive business strategy.
How Do You Fit AI Governance into an Existing Data Governance Program?
As enterprises race to adopt new technologies, one of the most hotly debated questions among data leaders is how to properly integrate artificial intelligence governance into established data governance frameworks.
Nicola Askham has observed that existing data governance teams are increasingly being asked to absorb heavy AI governance responsibilities, frequently without receiving any additional resources or the specialized technical skills that AI governance uniquely demands. While data governance professionals often feel comfortable managing the data quality components of AI, many feel entirely unequipped to handle complex tasks such as rigorous model validation or screening for algorithmic bias. Despite feeling overwhelmed, many practitioners take on the challenge simply because they believe it is better managed by them than ignored altogether.
At LMCU, Kevin Guyot’s approach involves consolidating governance councils wherever feasible and establishing a rigorous vetting process before any AI use case is permitted to move forward. These evaluations go far beyond basic technical feasibility, forcing teams to confront fundamental questions: Does it genuinely make sense to deploy AI for this specific capacity? If so, does the organization possess data of sufficient quality to allow the system to act reliably? And ultimately, just because artificial intelligence can be deployed in a certain way, should it be? This approach ensures that all AI initiatives remain firmly aligned with the organization’s core values, ethical standards, and guiding principles.
David Woods noted that hundreds of customer conversations reveal a common pitfall: many organizations initially approached AI governance by simply compiling a list of approved software tools and establishing a basic use case evaluation process, while completely overlooking the underlying data requirements until it was too late. Woods emphasizes that existing governance frameworks already contain vital elements like context, lineage, and quality rules, which represent precisely what AI models require to function accurately. The key is to leverage those existing foundations while recognizing that AI introduces entirely new considerations, including new personas and roles spanning legal, ethics, and human resources departments that historically did not participate in traditional governance councils.
Furthermore, research underscores that approximately nine out of ten AI use cases still fail to reach full production status, with root causes almost invariably tied back to data deficiencies. Enterprises that successfully transition AI pilots into production are typically those that recognize early on that governance is not a bureaucratic gate standing in the way of deployment, but rather the essential foundation supporting the entire structure.
Where to Start with Data Governance for AI
When considering strategic priorities for the months ahead, industry experts stress the importance of focusing on foundational mechanics rather than chasing superficial trends. Kevin Guyot reiterates that robust governance must remain the mandatory starting point, as it provides the critical business context required for AI models to consume information reliably and minimize hallucinations. Without establishing common organizational terminology and trustworthy data quality, achieving true AI readiness remains impossible.
Nicola Askham presses the point that corporate investment must accurately reflect organizational priorities. In too many instances, corporate leaders have made data governance personnel redundant as part of broader AI investment pushes, fundamentally misunderstanding that data governance and artificial intelligence are not competing priorities, but mutually dependent pillars.
Meanwhile, David Woods recommends that organizations focus heavily on refining their business semantic layer—the contextual framework that grants AI systems a coherent, governed understanding of what enterprise data actually means. Ensuring that this contextualized semantic layer is fully accessible not just to human business and IT users, but crucially to autonomous agentic systems, represents a major competitive differentiator for modern enterprises.
Ultimately, the consensus among industry leaders indicates that data governance is no longer a peripheral program running parallel to an artificial intelligence strategy; it is the operational substrate upon which successful AI initiatives operate. Organizations making genuine progress are those building governance and strategy in tandem, embedding them directly into everyday workflows, and continuously evaluating whether their underlying data is robust enough to support autonomous action.
