Interview with Keith Schlosser
Strong data foundations, not just new tools, determine whether enterprise AI and modernization succeed or fail in complex organizations.
Keith Schlosser, a multi-time CIO, discusses how foundational data and content quality determine the success or failure of enterprise modernization and AI initiatives, drawing on decades of experience leading large-scale transformation in complex, regulated environments. He highlights how organizations often underestimate the importance of structure, governance, and context in their information systems, leading to stalled or underperforming modernization efforts. His perspective emphasizes that sustainable AI adoption depends on getting these fundamentals right before scaling technology across the enterprise.
To begin, what experiences in your early career shaped how you now approach leadership in technology and large-scale change?
Early in my career, I watched the customer relationship management (CRM) software wave unfold. Every enterprise rushed in—massive investments, aggressive timelines, and enormous expectations. And many of those initiatives underdelivered. Not because CRM was flawed, but because the underlying data was.
That lesson has stayed with me for 36 years: technology doesn’t fail—foundations do.
Real success comes from preparing the organization—getting your content, data, and governance in order before you scale. That’s not always popular, but it’s what works.
We are seeing a similar pattern with AI. Many organizations are trying to move quickly by stitching together tools themselves or relying on AI features bolted onto existing systems. That can work for early experimentation, but it breaks down when you try to apply AI to real business processes. The issue is not simply whether you have access to data or content. It is whether that information has the structure, context, governance, and quality needed for AI to use it reliably. But with AI, the stakes are even higher. If your content and data are fragmented, poorly governed, or lack context, AI will simply scale those problems faster and more expensively.
As a leader, you have to be willing to push back on “move fast at all costs.” Real success comes from preparing the organization—getting your content, data, and governance in order before you scale. That’s not always popular, but it’s what works.
That does not mean every organization has to manually prepare everything before it can start. With the right platform, much of that work can be automated. But if you take a DIY approach, or assume that AI layered onto legacy systems will solve the problem on its own, you quickly run into the same foundational issues that hurt earlier waves of enterprise software. Real success comes from building on a foundation that can prepare content, apply governance, preserve context, and scale responsibly.
As your career progressed into leading technology and insurance organizations, what key lessons have you learned about helping large companies successfully manage change and modernisation?
Modernization efforts fail for predictable reasons—and most of them are self-inflicted.
Large organizations consistently try to modernise on top of content environments they don’t fully understand. Legacy enterprise content management (ECM) platforms were meant to address this, but it never fully delivered on the promise of a clean, governed, enterprise-wide source of truth. In most organisations, content remains scattered across systems, duplicated across applications, buried in repositories, and stripped of the context needed to make it useful. Yet companies treat them as stable foundations.
They’re not.
If you don’t understand your content, you can’t modernise it. And in an AI context, that becomes a hard stop. AI depends on context, structure, and quality—which rarely exist in most legacy environments.
The hard truth is simple: you cannot modernise on a broken content foundation. The organisations that succeed are the ones willing to address that reality directly rather than work around it.
As a next step, how are you seeing AI change the way companies manage and organise their documents and information today?
Most companies are still thinking about this backwards.
They are taking legacy ECM systems—designed for storage—and trying to layer AI on top. That approach will deliver incremental improvement at best and disappointment at scale.
AI changes the role of document management entirely. It’s no longer about storing and retrieving files or moving everything into one central repository—it’s about understanding content and making it usable by AI-driven workflows in real time.
An AI-native content platform treats content as structured, contextualised data from the start. It creates an intelligent layer that can ingest content from across the enterprise, preserve its original context, enrich it with structure and metadata, and make it usable for AI without forcing a wholesale migration. This enables capabilities like Retrieval-Augmented Generation (RAG), where AI models can securely access and use enterprise content with accuracy and relevance.
Without that foundation, RAG doesn’t work well. You get hallucinations, inconsistent answers, and low trust. With it, you get something very different—reliable, explainable AI grounded in your actual business content.
The issue is that many leaders have accepted that ECM is inherently clumsy and disappointing. That assumption is becoming outdated.
So yes, we are seeing a different way emerge. It is AI-native content infrastructure that prepares information as it enters the platform, connects it to business context, and allows organizations to use their content as the foundation for trusted AI.
Building on that, where do you see the biggest opportunities for AI to take on more complex work in document and information management that is still not being fully explored?
The biggest opportunity is moving beyond storage into true content understanding and activation.
Most ECM systems today are still glorified repositories with search layered on top. That’s a low-value use of very high-value information.
AI changes that. With an AI-native approach, you can:
- Understand documents at scale—context, meaning, relationships
- Extract and structure key data automatically
- Feed that content into RAG pipelines for accurate, context-aware AI outputs
- Drive intelligent workflows based on what the content actually says, not just where it’s stored
There’s also a major, often overlooked opportunity in migration. AI can dramatically reduce the cost and effort of moving off legacy platforms by classifying, rationalising, and enriching content during the transition.
The reason this isn’t happening broadly is simple: many organisations know how long, expensive, and cumbersome ECM-to-ECM migrations are, and they have become conditioned to expect very little from their ECM stack. That mindset is now the biggest barrier.
Turning to leadership, what does effective leadership look like today when teams are working with fast-changing technology while still needing trust and stability?
Leadership today requires clarity and restraint as much as vision.
There’s a lot of noise around AI, and teams can easily get pulled in too many directions. Strong leaders focus on what actually matters: governance, security, and foundational integrity.
Strong leaders focus on what actually matters: governance, security, and foundational integrity.
In document and content management, governance and security are always top of mind, and while legacy ECM systems are well governed and secure, they’re expensive to maintain, difficult to upgrade, and simply aren’t architected to leverage AI effectively.
This is especially critical as we move into compliance-aware AI—systems that not only generate insights but do so with auditability, explainability, and regulatory alignment built in.
Chasing the latest tool without addressing those fundamentals creates risk, not progress. Stability comes from discipline, not from slowing down innovation, but from directing it properly.
In your view, what separates leaders who are adapting well to these changes from those who are finding it difficult?
It’s not a technology gap—it’s a mindset gap.
The leaders who are succeeding understand a few things very clearly:
- AI will not fix bad data or poor content—it will expose it
- A test-and-learn approach is essential, but it has to be grounded in reality
- Content and data foundations are strategic assets, not operational afterthoughts
- There is a fundamental difference between bolting AI onto legacy platforms and adopting AI-native architectures
The ones struggling are often trying to protect past investments or extend platforms that were never designed for this era.
At some point, you have to be willing to say: this approach has run its course.
Finally, as you think about the next 5 to 10 years, how do you see AI changing the way companies operate, especially in regulated industries?
AI will redefine how regulated industries operate—but not evenly.
The winners will be the organisations that invest in becoming AI-native at the content layer. That means treating documents not as static records, but as dynamic, governed, and fully understood assets.
In practical terms, that enables:
- Real-time decisioning grounded in enterprise content via RAG
- End-to-end auditability through compliance-aware AI
- Faster response to regulatory change
- Significantly improved operational agility
The laggards will be those that continue to rely on bolt-on approaches and legacy ECM platforms that were never designed for this level of intelligence.
This isn’t about incremental improvement—it’s a structural shift.
And like every major shift I’ve seen over the past 36 years, the organisations that focus on the fundamentals early will be the ones that lead when the market catches up.
Keith Schlosser
