Enterprise AI stalls not because models fall short, but because years of neglected data governance leave organizations unable to scale intelligence safely or consistently.
Kjersten Moody, 3x CDO and CDAO In Residence at Insight Partners, draws on her experience across large enterprises to show how legacy teams are forced into costly data retrofits while AI-native companies move faster by design.
The path forward embeds metadata, lineage, ownership, and governance into products and operating models from day one, treating clear constraints as the mechanism that enables speed rather than blocks it.
The enterprise AI conversation has become fixated on model sophistication, but the algorithm isn't the constraint. It’s the accumulated weight of data governance debt. As organizations race to deploy intelligent systems, many are discovering they can’t move forward because the foundations aren’t there. Metadata, data lineage, data quality, ownership, and architecture are the structural prerequisites that determine whether AI can work at scale at all.
Kjersten Moody has seen this pattern play out from the inside. A three-time Chief Data Officer and now CDAO In Residence at Insight Partners for Enterprise, she previously served as the inaugural Global CDO at Prudential Financial and the inaugural CDAO at State Farm. Across roles that include managing more than $100M in budgets and leading global teams of over 1,000, Moody has formed a clear perspective: many organizations pointing to early AI wins aren’t breaking new ground. They’re finally reaping the returns on foundational data work put in place years earlier.
"The principles of data management and governance are more mission-critical today than they have ever been. Concepts like metadata, data lineage, data quality, data ownership, and data architectures are the key ingredients to having your data be AI-ready," says Moody. "That’s not a hype cycle; that’s just a fundamental truth about how AI is constructed on both its generative and predictive sides."
The retrofit challenge: The gap between foundational readiness and AI ambition creates a divide between incumbents and disruptors. For many a substantial legacy organizations, the primary challenge is what Moody calls a "retrofit" to pay down generations of legacy data debt. "In the longstanding enterprise, the strategic challenge for data is that of a retrofit, where you must take multiple generations of data from pre-AI systems and find a way to make it fit for purpose for AI at scale," she explains. "That is a very difficult proposition."
The native advantage: The proactive approach—building adaptive foundations designed for intelligence from inception—moves beyond the reactive retrofit work common among incumbents. "Newer, venture-backed, AI-native startups have a huge advantage because they don't have this legacy data debt," she continues. "They have the opportunity to begin with the current end in mind, and it is an absolute competitive advantage."
Data Dewey Decimal: "The strategy is to build the principles of well-managed data into the software engineering and product processes from day one. Your data is in a catalog, your metadata exists and is available, and there are clear design patterns for fitting third-party data into the ecosystem. It's like setting up a library where you already know the design pattern; for every new book you put on the shelf, you understand how to place it so it can be fully leveraged. It becomes not a data officer request, but a core skill where software engineers, business users, and AI engineers all have a component of healthy data built into their job mission. That way, you never have to retrofit. Instead of retrofitting, you're evolving."
Moody acknowledges the heavy lift this foundational work requires, framing it as a direct response to the urgent, board-level pressure for AI solutions many leaders face today. Framing it this way connects the unglamorous work of data management directly to business value.
Paying dividends: "It’s a very heavy lift to install the tools and operating model for well-managed and governed data. But the dividends are clearly shown now with the AI demands that businesses and boards are asking," Moody says.
The conversation also reframes governance—often seen as a roadblock—as a prerequisite for speed. Because a lack of formal rules can lead to risky "shadow AI," a clear frame of reference—grounded in standards like the ISO's AI management system and the NIST AI Risk Management Framework—is what helps give teams the confidence to move fast. According to Moody, the cultural change required to embed this discipline is more achievable when the right leadership and structures are in place.
The need for speed: "Cars have brakes so they can go fast. Imagine how fast you would drive if your car had no brakes. That’s the frame of reference for governance. The frame's job is to stop you if the solution being undertaken is in violation of the law, regulation, or the company's principles. Within that frame, your job is to go as fast as possible. It is then leadership's obligation to provision the tools, capabilities, and upskilling that allow your people to deliver to their best ability," Moody explains.
Where there's a will: Culture doesn’t change through persuasion alone. It changes when expectations are clear, supported, and made part of how work actually gets done. "If leadership is bought in, the tools and design patterns are well-published, and the incentives are in place, the cultural change will be much easier," she notes. "If there's organizational will that is well-provisioned and incentivized, it will get done."
From where Moody sits today, the path forward is less about prediction and more about preparation. "If you're an up-and-coming leader, part of your job is to look around the corner, understand what's coming, and get your teams and company ready to embrace it. You need to be tapped in and building strong relationships with the innovation community, because they are driving that next-generation paradigm shift for the delivery of data and AI in the enterprise, full stop," she concludes.
