A few years ago, a Knowledge Management team I am aware of spent considerable effort documenting the lessons from a major programme. They interviewed people, captured the key insights, organised them clearly and stored everything in a system that was genuinely accessible.

Six months later, the organisation launched a similar programme.

Almost none of the captured lessons informed the new decisions being made.

Not because the knowledge was hidden. Not because people were indifferent. But because there was no real connection between what had been captured and the moment when decisions were actually being made. The knowledge existed in one place. The decisions happened somewhere else.

Capture is not the same as use

Knowledge Management has made genuine progress over the past two decades. Organisations capture more than they used to. They share more. They structure experience into frameworks, lessons and repositories that previous generations did not have.

But capturing knowledge and ensuring it reaches decisions are two different things.

A lessons-learned repository is not, by itself, a decision input. An expertise directory does not automatically put the right knowledge in front of the right person at the right moment. A documented best practice does not guarantee it shaped the choice that was made.

The path between what an organisation knows and what actually influences its decisions is often shorter in theory than in practice.

Why this gap is structural, not a failure

It is worth being precise about this, because Knowledge Management professionals sometimes hear this observation as a criticism. It is not.

KM was largely designed to solve a real and important problem: organisations were losing knowledge they needed to retain. The response — capture it, structure it, share it — was the right one.

But there is a second problem that sits adjacent to the first: even when knowledge is captured, it may not reach the decisions it was meant to support.

That is not a KM failure. It is a structural gap between two things that are related but not automatically connected — knowledge management and decision-making.

Closing that gap requires something more deliberate than a better repository or a more comprehensive lessons-learned process. It requires thinking about the path from knowledge to decision as something worth designing.

What that connection looks like in practice

It does not need to be complicated.

For an important decision, it is worth asking explicitly: what does the organisation already know that is relevant here? What lessons exist? What expertise is available? What have we learned from similar situations?

And then — crucially — making sure the answers actually reach the people making the decision, at the moment they are making it, in a form they can use.

This is different from storing knowledge well. It is about ensuring that knowledge travels to the right moment.

It also means preserving, after the decision, what knowledge was used and how it shaped the reasoning. Because that connection — between knowledge and decision — is exactly what tends to disappear over time, as last week's article explored.

Why this matters now

Decisions are where knowledge either proves its value or quietly fails to.

An organisation can have an excellent KM function and still make decisions that ignore what it knows. It can have well-maintained repositories and still repeat avoidable mistakes. Not because KM is not working, but because the link between knowledge and decision was never explicitly made.

Structured Decision Continuity examines that link — not as a criticism of Knowledge Management, but as an extension of what KM can offer when it is connected more deliberately to the decisions it is meant to support.

The question worth asking is not only: do we capture our knowledge?

It is: does our knowledge reach our decisions?

‍

This article is part of a monthly series on how organisations can preserve the reasoning behind important decisions, explored through the lens of Structured Decision Continuity.

___________________________________

We talk a lot about KM implementation. But how many of us have stopped to ask — is our KM framework upgraded and evolved for the AI era?

I created a KM roadmap a while back to help KM folks like me have a structured approach to knowledge management. The six steps — from defining objectives to measuring outcomes — remain as relevant as ever. But times have changed, and so has our KM roadmap.

AI is no longer a future consideration. And if our KM systems are not designed with AI in mind, we are leaving enormous value on the table.

So I went back to my original framework and asked one simple question at every step: where can AI make this smarter, faster, and more impactful?

Here is what that looks like:

When you define objectives, AI can analyse patterns across your organisation to predict which knowledge gaps are causing the most friction — before your customers even tell you.

When you identify knowledge sources, AI can crawl across your systems, documents, and conversations to surface the knowledge that already exists but nobody can find.

When you choose your KMS, look beyond traditional systems. AI-native platforms with smart search, auto-tagging, and content recommendations are now the baseline, not the premium.

When you design your KM plan, let AI do the heavy lifting on categorisation, taxonomy suggestions, and flagging content that has gone stale or outdated.

When you train for cultural shift, AI can create personalised learning paths so every team member gets the knowledge most relevant to their role — not a one-size-fits-all training deck.

When you measure and evaluate, AI dashboards can track not just knowledge usage but also real CX outcomes — CSAT, first-contact resolution, average handling time — connecting your KM investment directly to business results.

This is not about replacing the human side of knowledge management. It is about amplifying it by using AI as your assistant..

Your people still drive the culture. Your experts still create the insight. AI simply helps you do more with what you already have by giving you time to focus on what matters and freeing up your time for things that you can automate.

If you are a KM professional thinking about where to focus your energy this year, start here. Not by overhauling everything — but by adding the AI layer, one step at a time.

I would love to know — which of these six steps do you think AI can impact the most in your organisation? Drop your thoughts in the comments.

_______________________________

Artificial intelligence is reshaping how organizations handle information and influence decisions. Many treat it as a replacement for Knowledge Management, assuming better answers will follow.

The real challenge is how people think, question, and decide together with AI, which makes Conversational Leadership a practical discipline for responsible judgement and action.

‍

Artificial intelligence is reshaping how organizations handle information and what we often call knowledge. It is tempting to see it as a replacement for Knowledge Management, a more capable system that finally delivers what earlier approaches struggled to achieve. In one sense, that is understandable. AI can capture, retrieve, and synthesize information at a scale and speed that traditional repositories, taxonomies, and search tools never managed.

But if that is all we mean by Knowledge Management, then we have reduced it to something quite limited.

The deeper ambition was never just better storage or faster access. It was always about better judgment, better learning, and better decisions in situations that are often messy and uncertain. The challenge was never simply information. It was how we make sense of it together.

AI changes the terrain. It does not just store or retrieve information; it can participate in our flow of thinking. It can reframe questions, suggest connections, and influence what we notice. When we begin to think with AI rather than only use it as a tool, the line between information and knowledge becomes less clear.

AI works with representations of the past. It does not experience the present as we do, and it does not bear responsibility for what follows. That remains with us.

This matters because AI outputs often feel fluent and convincing. The risk is not that we know too little, but that we accept too quickly. We may find ourselves agreeing without fully examining what is being suggested or overlooking what is missing.

As AI strengthens the informational backbone of organizations, the real work shifts. It moves toward interpretation, alignment, and responsible action. It asks more of us in how we question what we see, how we surface assumptions, and how willing we are to stay with uncertainty rather than close things down too quickly.

Conversation becomes central here, but not just any conversation. Many organizational conversations reinforce existing patterns, avoid challenge, or defer to authority. For conversation to be useful in this context, the conditions need to support curiosity, allow for doubt, and enable thinking things through together without rushing to agreement.

This is where Conversational Leadership comes in, not as a role or a position, but as a practice. It is about creating the conditions in which people can think together more carefully, especially when the issues are complex and the answers are not obvious.

In the age of AI, that practice extends to how we engage with the technology itself. If AI becomes part of how thinking happens in organizations, then it also becomes part of the conversation. It needs to be questioned, tested, and worked with, not simply accepted.

Seen this way, AI is not an oracle that provides answers, but a participant in a broader system of sense making. It can extend our thinking, but it does not replace our responsibility for judgment, ethics, or action.

So, the question is less about what AI can do, and more about how we respond to it. Knowledge Management, in this light, becomes less about systems and more about our collective ability to make sense of things together in environments where AI is always present.

The tools will continue to evolve. The need to think well together, and to take responsibility for what we decide and do, remains a human concern.

____________________________________

‍

In many discussions about AI literacy, a natural follow-up question quickly appears: What does Knowledge Management literacy mean inside organisations?

Τhe term is often mentioned, but rarely explained in practical terms.
‍
Knowledge Management literacy is not primarily about tools or platforms.
‍

It is about understanding how organisational knowledge is recognised, structured, and validated so that it can reliably support decisions.The simple framework below summarises four practical capabilities that shape how organisations work with knowledge.
‍

These capabilities may appear straightforward. In practice, they often determine whether knowledge supports sound decisions or simply turns into fragmented information.

1. Locate Knowledge

‍The first capability is the ability to locate where organisational knowledge actually lives.
‍

‍

In many organisations, knowledge is distributed across multiple systems: document repositories, collaboration platforms, shared drives, internal portals, and email archives. Without a clear understanding of this landscape, people often spend a considerable amount of time simply searching for information.

Knowledge Management literacy therefore begins with a basic awareness of the organisation’s knowledge environment: where different types of knowledge are stored and which systems serve which purpose.

Without this basic capability, organisations struggle to use knowledge consistently, whether by people or by AI systems.

2. Identify the Authoritative Source

‍Locating information is not enough. The next step is recognising which version of knowledge can be trusted.

‍

‍

In practice, organisations often operate with multiple versions of the same document, guideline, or procedure. Teams may rely on different sources without knowing which version is officially maintained.

Knowledge Management literacy therefore includes the ability to identify the authoritative source: the version of knowledge that is validated, maintained, and intended to guide decisions.

‍3. Understand Knowledge Context‍

Knowledge is never created isolation. It always emergies in a particular context: a regulatory environment, a project phase, and a specific organization challenge.

‍

‍

Understanding this context is essential for interpreting knowledge correctly. Without it, documents and guidance may easily be reused in situations where they no longer apply. Knowledge Management literacy therefore involves recognising how and why knowledge was produced, and under which conditions it should be interpreted.

4. Validate Knowledge before reuse

Finally, knowledge must be validated before it is reused, shared, or embedded in automated processes.

‍

‍

Organisations evolve, policies change, and procedures are updated. If knowledge is reused without verification, outdated information can easily spread across teams or systems.

Knowledge Management literacy therefore requires the ability to confirm that knowledge remains current and relevant before it is applied again.

Why these capabilities matter for AI

These four capabilities become particularly important as organisations explore AI-enabled systems.

AI can retrieve, process, and connect information at scale. However, the quality of its outputs depends directly on the structure and reliability of the knowledge it accesses.

If knowledge sources are fragmented, unclear, or outdated, AI may simply accelerate confusion rather than support judgement.

For this reason, developing Knowledge Management literacy is not only a Knowledge Management concern. It is increasingly becoming a foundational capability for organisations seeking to use AI responsibly and effectively.

Future Knowledge Nuggets will explore these capabilities in greater detail and examine how organisations can strengthen them in practice.

‍

Disclaimer: The views expressed in this article are my own and do not represent the position of my employer or any institution I am associated with.

__________________

In the modern enterprise landscape, knowledge bases are increasingly shaped by employees and customers rather than by vetted internal experts alone. While this democratization of sharing information has reached new heights in volume, depth and trust, it also introduces significant management challenges for organizations.

‍

As the line between verified data and informal reviews blurs, companies must develop robust governance structures to ensure user-generated content remains a net positive. This ensures enterprises can reap the benefits of open communication and ownership without the risk of misinformation spread and reputational damage.

Identifying the Risks of Unmoderated Knowledge

By understanding the key risks associated with user-generated content in an enterprise setting, institutions can effectively develop appropriate moderation protocols to address them.

Inaccurate Content

Inaccurate information often spreads quickly when platforms lack the proper validation protocols. Employees might assume that the information on a shared document is accurate without double-checking whether the contributor has uploaded outdated procedures or incorrect technical specifications. This could lead to false information cascading throughout the organization, leading to costly mistakes and a decline in trust regarding the central repository.

Lacking proper moderation protocols also leads to an influx of informal tips. While not overtly false, these entities rarely undergo the scrutiny required for professional standards. An accumulation of unverified entries results in a lack of cohesion, making it difficult for knowledge management professionals to find a single source of truth. This concern highlights the importance of verification methods for accuracy and consistency.

Algorithmic and Human Bias

User-generated contributions often contain mild biases, though they may be unintentional. A lack of neutrality slowly morphs the entire knowledge ecosystem. In large enterprises, this could result in departmental silos that favor worker preference over efficiency. Such tendencies can hinder collaboration and prevent the organization from scaling its knowledge effectively across teams.

Additionally, search algorithms may prioritize engagement over the accuracy of the information they share. This creates an environment where popularity triumphs truth, resulting in flawed information remaining visible because it’s frequently accessed. To ensure that engagement-driven content doesn’t overshadow reliable data, management teams should build digital systems where accuracy dictates visibility.

Operational Friction

Massive quantities of unmanaged content also mean employees spend more time and energy finding the answers they need. This friction increases staff members’ cognitive load and can lead to abandonment of collaborative tools. Without an ergonomic way to infer key information for day-to-day operations, efficiency inevitably drops.

Furthermore, operational friction creates onboarding complications. New employees have more difficulty filtering through the noise of unverified user-generated content, leading to confusion and operational inefficiencies. This challenge underscores the importance of proactive content management to ensure a streamlined user experience.

Legal and Reputational Damage

Internal knowledge bases must comply with key regulations, especially when handling large volumes of sensitive data. While catastrophic data breaches from sophisticated cyber attacks are common today, poor internal handling is also a prominent cause of leaks. Allowing exchanges to go unmonitored means that protected information circulates too freely. A lack of oversight could be detrimental to a business’s legal standing.

The long-term impact on a company’s image is a greater threat. This is a difficult area to navigate because digital content creates unique challenges for reputation management, where a single unvetted post can compromise stakeholder trust. Proactive moderation is a fundamental tool for protecting a brand’s perception and stability.

Building a Strong Governance Framework

Establishing meticulous verification procedures is key to mitigating the operational and financial risks posed by user-generated content in an enterprise setting.

Technical Moderation

Automated workflows can be incredibly efficient at flagging noncompliant or inaccurate content before more people view it. However, technical information should require human expertise to verify in its context. In general, having a tiered verification system allows content entering the knowledge base to receive adequate attention depending on its importance.

Moderation processes can be further improved by leveraging metadata. In an internal knowledge base, expiration dates and version control prevent the accumulation of outdated content. When systems automatically prompt users to remove or archive content as its expiration date approaches, the repository can remain uncluttered and high-quality. This approach also reduces the burden of manual oversight.

Fostering a Culture of Responsible Creation

Technology and policy require a strong foundation in organizational culture to be truly effective. Employees should be trained to understand the importance of ethical and efficient information distribution.

By ensuring that staff members are deeply aware of key regulations and frameworks, organizations can be confident that their knowledge base stays compliant and genuinely valuable to their employees.

Keeping Enterprise Knowledge Bases Efficient and Valuable

Institutions that have strong governance over their knowledge bases are providing significant benefits to their employees, ensuring that all internal information they encounter is accurate and genuinely helpful. Yet it is also important that enterprises strike a balance between vigilant oversight and open communication, enabling team members to foster a sense of ownership and authority. An investment in employees can support long-term company resilience

___________________________________________