(This post originally appeared on my LinkedIn Profile.)
A colleague said something to me recently that gave me pause.
“You’re a bit of a Luddite when it comes to AI, aren’t you?”
It was said with a smile, the kind of gentle ribbing that comes from genuine curiosity rather than criticism. But it stuck with me, because I realised I haven’t been clear enough about where I actually stand.
So let me be direct: I think AI is remarkable. I use it. I find it genuinely useful. I’m curious (and also deeply concerned) about where it’s headed. And in the context of Knowledge Management specifically, I believe it has real and growing value.
What I don’t believe, and what I’ll continue to push back on, is the idea that AI solves Knowledge Management. Because that framing, however appealing, leads organisations down a very expensive and frustrating path.
The Seductive Logic of the Tech Solution
There’s a pattern I’ve watched repeat itself across organisations for years, and AI is simply the latest chapter in a long story.
It went like this: first, we were going to solve knowledge management with intranets. Then wikis. Then SharePoint. Then enterprise social networks. Then big data platforms. Each time, the pitch was essentially the same: “This technology will finally make your organisational knowledge accessible, searchable, and usable.”
And each time, organisations discovered the same uncomfortable truth: the technology worked. It was the knowledge management that didn’t.
Files sat untagged. Content became stale within months. People reverted to emailing each other, because the system was technically available but practically ignored. Adoption curves flattened. The investment quietly became shelfware.
AI doesn’t change this dynamic. It accelerates it, in both directions. Deploy AI on top of well-structured, well-governed, actively maintained knowledge assets, and it can be genuinely transformative. Deploy it on top of the organisational equivalent of a cluttered attic, and it will confidently surface the wrong answer, twice as fast. (Wrong answers at scale, as I recently commented to another colleague.)
Knowledge Management Is a People Problem First
Here’s the thing that no vendor wants to put on a slide: most knowledge management failures are not technology failures.
They’re culture failures. Process failures. Leadership failures.
When an organisation loses critical expertise because a senior employee retires, that’s not a software gap, it’s a knowledge transfer gap that nobody prioritised until it was too late. When two teams duplicate months of research because they had no way of knowing what each other had already done, that’s a collaboration and communication failure. When a new hire spends their first three months reinventing wheels, that’s an onboarding and documentation failure.
No AI tool, however sophisticated, fixes those problems at the root. It might paper over some of the symptoms and sometimes that’s genuinely useful, but it doesn’t address the underlying conditions that created the problem in the first place.
The Five Pillars That Actually Matter
Effective Knowledge Management rests on five interconnected foundations:
People: Knowledge lives in people’s heads first. Building a knowledge-sharing culture, incentivising contribution, and reducing the friction between “someone knows this” and “the organisation benefits from this” is fundamentally a human challenge. It requires trust, psychological safety, and genuine leadership commitment.
Process: How does knowledge get created, validated, stored, retrieved, and retired? Without intentional, well-designed processes, even the best technology becomes a dumping ground. Process is the architecture that gives technology its shape.
Technology: Yes, absolutely. The right tools, well-chosen and properly implemented, are a genuine enabler. AI included. But technology is an enabler of a good strategy, not a substitute for one.
Governance: Who owns the knowledge? Who decides what’s accurate, current, and trusted? Who has authority to retire outdated content? Without governance, knowledge systems decay. This is the unglamorous work that makes everything else sustainable.
Strategy: What are you actually trying to achieve? Knowledge Management without a clear strategic anchor is just tidying. It needs to connect to organisational goals: accelerating innovation, reducing risk, improving customer outcomes, developing capability. Strategy is what makes the whole system purposeful.
These five pillars don’t operate in isolation, they’re deeply interdependent. A brilliant technology choice undermined by weak governance will still fail. Strong processes with no cultural buy-in will still fail. Strategy without execution infrastructure will still fail.
There Are No Easy Answers and That’s Important to Say Out Loud
One of the most valuable things a Knowledge Management professional can do is resist the pressure to promise simplicity.
Organisations want easy answers. They want to hear that the new platform, the new AI integration, the new taxonomy framework will sort things out. And sometimes, in the short term, a well-chosen tool can create enough momentum to feel like progress.
But sustainable Knowledge Management, the kind that actually compounds in value over time, requires patience, iteration, and a willingness to do genuinely hard organisational work. It requires leaders who understand that knowledge is a strategic asset, not an IT problem. It requires investment in people and process, not just licences and implementations.
The organisations that get this right don’t treat Knowledge Management as a project with an end date. They treat it as an ongoing capability, like finance or HR, something that needs continuous attention, resourcing, and refinement.
So Where Does That Leave AI?
Exactly where good tools always belong as a powerful enabler within a thoughtful, well-structured Knowledge Management strategy.
Used well, AI can dramatically improve knowledge discovery. It can surface connections between documents that no human would have the bandwidth to find. It can reduce the cost of creating knowledge assets. It can personalise the knowledge experience for different roles and contexts. These are real benefits, and they matter.
But AI doesn’t replace the need for governance. It doesn’t create a culture of sharing where none exists. It doesn’t write your knowledge strategy for you, or decide which expertise is critical to capture, or ensure that the knowledge your organisation produces is accurate and trustworthy.
Those are human responsibilities. And until we’re honest about that, we’ll keep buying technology solutions to organisational problems and wondering why the needle doesn’t move.
I’m not anti-technology. I’m anti-magical-thinking.
The organisations that thrive in the long run are the ones willing to do both: invest in excellent tools and do the harder, slower, more human work of building the conditions in which those tools can actually deliver.
That’s not a pessimistic message. I think it’s actually a hopeful one. Because it means the answer is within reach, it just takes more than a software purchase to get there.