AI can make expertise more valuable. But only if it’s first made visible, usable and ready to travel.
Nine years ago, the internal knowledge team I worked in was made redundant.
IT had suggested to Finance that AI would soon be able to do much of what we did: expert knowledge transfer, lessons learned, onboarding support, internal good practice, and helping one part of the business reuse what another part had already worked out.
The assumption was that technology would make that work automatic.
It was a painful moment. Even then, though, it was clear that AI would change how organisations used internal knowledge. Since then, AI hasn’t removed the need for excellent knowledge work. It’s made the gaps easier to see.
We still hear versions of the same assumption today.
“AI will find it.”
“AI will summarise it.”
“AI will turn it into training.”
“AI will help people access what they need.”
Sometimes that’s true. But only if the knowledge is there, worth finding, and structured well enough to be reused.
AI doesn’t magically fix poor knowledge retention practices. In many cases, it exposes them: obsolete documents treated as sources of truth, missing material no one can find, and generic output created because the real expertise is still in people’s heads.
The issue isn’t really about your AI tool
Organisations are right to manage AI carefully. Platforms, policies and data controls shape what people can safely use.
But senior leaders shouldn’t assume the AI environment will stay neat, stable or fully controlled. Tools will change. Models will change. People will experiment. New use cases will appear faster than governance can manage.
So the long-term question isn’t just, “Which AI tool should we use?”
It’s also, “What knowledge is valuable enough to be used well, whatever the tool?”
Many organisations are weaker on this than they think.
They have documents, recordings, slide decks, SharePoint folders, project archives and handover notes. But that’s not the same as usable expertise.
Usable expertise includes context and judgement. It explains why a decision was made, what was tried before, what failed, who was involved, what the warning signs are, and what someone with experience would pay attention to next time.
AI can make that knowledge easier to search, reshape and activate. It may even create confident answers from weak material. But it can’t recreate expertise the organisation never captured.
AI-ready doesn’t mean AI-mandatory
Making knowledge AI-ready doesn’t mean forcing it into a particular AI tool, model or platform. It doesn’t mean every output needs to become a chatbot, automated guide or digital assistant.
AI-ready means the knowledge has been captured and structured clearly enough to be reused by people, teams, learning designers, search tools or future AI applications that may not exist yet.
The AI landscape inside organisations will keep changing. Well-captured knowledge still holds its value. If it’s clear, contextual and usable, it can travel across tools, teams and moments of need.
Start with your pockets of brilliance
The work organisations want AI to support isn’t new: expert moves, major project learning, after-action reviews, onboarding, and taking a solution from one part of the organisation into another.
Most large organisations already have pockets of brilliance.
A team that solves a recurring problem better than everyone else. A site that has found a smarter way to onboard new leaders. A function that handles customer escalations with less friction.
The real question is why that brilliance hasn’t travelled.
Is it because the knowledge sits in people’s heads? Is it buried in an old deck? Was it shared once in a meeting and then lost? Was the lesson captured, but never converted into something others could actually use?
If the knowledge isn’t visible, structured or usable, AI won’t fix that by itself.
It may find the document. It may summarise the transcript. It may produce a neat answer. But that doesn’t mean the organisation has incorporated the knowledge into day-to-day operations.
Where KC helps
At KC, this is one of the things we help clients do.
We help organisations capture critical knowledge, convert it into a usable format, and activate it where people need it.
It’s only one part of our wider work. Much of an organisation’s performance still depends on knowledge that’s valuable, vulnerable and often invisible.
Start with the parts of your organisation that excel. Ask why that knowledge hasn’t travelled further. Then ask what would need to be captured, structured and made usable before AI can help more people benefit from it.
Critical knowledge, made usable at the point of need.
If you’d like to talk about protecting your critical knowledge, we’re here.
