Is shadow learning the same as job shadowing?

No. Job shadowing means observing a colleague to learn their role, and it is a formal, scheduled practice. Shadow learning is unsanctioned and invisible: it is the self-directed learning employees do on their own, in tools their employer does not track.

How is shadow learning related to shadow AI?

They are the same behaviour seen from two departments. Shadow AI is IT's name for employees using unsanctioned AI tools; shadow learning is L&D's name for the learning that happens inside them. Microsoft and LinkedIn's 2024 Work Trend Index found 78% of AI users bring their own AI tools to work.

Is shadow learning a problem?

The learning is not the problem - the invisibility is. Shadow learning signals genuine, urgent demand that your catalog failed to meet. Treated as a compliance failure, it drives the behaviour further underground. Treated as a demand signal, it tells L&D precisely which skills the business needs right now.

How much of workplace learning is shadow learning?

Most of it. Scheduled, mandatory training - compliance modules, onboarding, certifications - covers roughly 5% of what employees actually learn at work. The remaining 95% is unplanned, triggered by a task, and happens outside any system L&D can see. That unrecorded majority is shadow learning.

The term

What is shadow learning?

Shadow learning is the learning employees do outside sanctioned systems - searching YouTube, asking ChatGPT, or messaging a colleague in Slack the moment a work problem appears. It is unrecorded, so L&D cannot see it, credit it, or build on it. In most organizations it is the majority of all workplace learning.

Last updated July 9, 2026 · by Eran Vaisfailr, Co-Founder & CEO of Plynn, previously L&D at Booking.com.

Some call this informal learning, or learning in the flow of work. Shadow learning names the specific mechanism underneath those broader terms: it is triggered by a real work signal, not general curiosity, and it is what employees reach for the moment sanctioned tools are too slow to answer.

Shadow learning vs. job shadowing: what's the difference?

The two are routinely confused, and they are opposites. Job shadowing is formal and scheduled: you observe a colleague in order to learn their role, and someone signed off on it. Shadow learning is unsanctioned and invisible: nobody scheduled it, nobody recorded it, and the only person who knows it happened is the employee who needed the answer.

Shadow learning is to L&D what shadow AI is to IT

IT already named this behaviour. When employees paste work into an unapproved AI tool, security calls it shadow AI, and it is one of the fastest-growing concerns in the enterprise. Microsoft and LinkedIn's 2024 Work Trend Index found that 78% of AI users bring their own AI tools to work.

Shadow learning is the same behaviour, viewed from the L&D side of the building. The employee who asks ChatGPT how to structure a QBR is doing both at once. IT sees a governance gap. L&D should see a curriculum request, submitted in the only format the employee had time for.

Why do employees learn outside the LMS?

Because the work cannot wait for a course. An LMS is built for scheduled, mandatory training - compliance modules, onboarding, certifications - and it is good at that. But that is roughly 5% of what people actually learn at work. The other 95% is unplanned and task-triggered, and it arrives with a deadline attached.

Josh Bersin named the destination in 2018: learning in the flow of work. Shadow learning is what employees do when nobody builds it for them (more on why employees don't use the LMS).

Where shadow learning starts: three signals

Shadow learning does not begin when someone opens a learning platform. It begins when work sends a signal, and the platform is not where the person is standing.

The signal	Where it fires	What L&D sees today
Someone prompts	Slack, Teams, ChatGPT, Claude	The question gets answered, and then it evaporates.
A task is assigned	Monday.com, Asana, Jira	The skill gap shows up as a slow ticket, not a learning request.
Feedback arrives	HiBob, Workday	A development note with no path attached to it.

How can L&D measure shadow learning?

You cannot instrument YouTube, and you should not try to instrument ChatGPT. Surveillance drives shadow learning further underground - the same lesson IT learned with shadow IT a decade ago. We go deeper on the mechanics in measuring informal learning.

The alternative is to meet the demand at the signal. If a prompt in Slack, a task on the board, or a line of feedback in your HRIS can each generate a real learning path in minutes, then the employee has no reason to go around you - and the request itself becomes the measurement. You stop counting course completions and start seeing what the workforce is actually trying to learn, on the day it tries to learn it.

That is what Plynn is built to do - showing up as a message in Slack or Teams, not a link to a course library. In our pilot, 60% of employees adopted it with zero nudges from L&D, and 90% came back for a second course - because the learning showed up where the work already was.

Frequently asked questions

What is shadow learning?: Shadow learning is the learning employees do outside sanctioned systems - searching YouTube, asking ChatGPT, or messaging a colleague in Slack the moment a work problem appears. It is unrecorded, so L&D cannot see it, credit it, or build on it. It is typically the majority of workplace learning.
Is shadow learning the same as job shadowing?: No. Job shadowing means observing a colleague to learn their role, and it is a formal, scheduled practice. Shadow learning is unsanctioned and invisible: it is the self-directed learning employees do on their own, in tools their employer does not track.
How is shadow learning related to shadow AI?: They are the same behaviour seen from two departments. Shadow AI is IT's name for employees using unsanctioned AI tools; shadow learning is L&D's name for the learning that happens inside them. Microsoft and LinkedIn's 2024 Work Trend Index found 78% of AI users bring their own AI tools to work.
Is shadow learning a problem?: The learning is not the problem - the invisibility is. Shadow learning signals genuine, urgent demand that your catalog failed to meet. Treated as a compliance failure, it drives the behaviour further underground. Treated as a demand signal, it tells L&D precisely which skills the business needs right now.
Why do employees learn outside the LMS?: Because the work cannot wait for a course. An LMS is built for scheduled, mandatory training - compliance, onboarding, certification. When a specific problem surfaces at 2pm on a Tuesday, employees go where the answer is fastest: search, AI, or the person one desk over.
How much of workplace learning is shadow learning?: Most of it. Scheduled, mandatory training - compliance modules, onboarding, certifications - covers roughly 5% of what employees actually learn at work. The remaining 95% is unplanned, triggered by a task, and happens outside any system L&D can see. That unrecorded majority is shadow learning.
How can L&D measure shadow learning?: You cannot measure it by instrumenting YouTube or ChatGPT. You measure it by meeting the demand where it starts - a prompt in Slack, a task on the board, feedback in your HRIS - and generating the learning there. The request itself becomes the measurement.

this loop already runs at your company. you just can't see it yet.

Book a Demo

Invite-only · SSO & SCIM included · Security review materials on request