1 July 2026
In-person AI training versus an LMS module
Completion is not capability. Monday morning is the test, not the LMS dashboard.
Who this is for
This is for the L&D or transformation lead who has been offered a recorded module, a self-paced LMS path, or a "digital academy" as the AI training plan. Completion rates will look tidy. The team will still work the old way on Monday, except for the two people who already taught themselves.
If the goal is a completion number, buy the module. If the goal is a different Tuesday, you need a room, a facilitator, and the team's actual work on the table.
The business problem
Self-paced content scales. That is why procurement likes it. It also leaves every hard question unanswered. What may go into this tool? What does a good check look like on our files? What does the manager do when the output is fluent and wrong? A recording cannot look at the learner's screen and say "stop, that customer record should not be in there."
McKinsey reported in 2025 that 88% of organisations use AI in at least one function while only about a third have scaled past pilots, and that only 39% see any EBIT impact. TalentLMS reported in 2025 that 49% of workers say AI is moving faster than their company's training. The Conference Board reported in 2026 that only 33% of workers received employer AI training in the last six months. An LMS module assigned to everyone is a common way those figures stay stuck: the organisation can say it assigned training, and the work does not change.
World Economic Forum reporting in 2025 found that 63% of employers name skill gaps as the top barrier to transformation. Skill gaps do not close because a video finished playing.
What an LMS module is good for
Be fair to the format. A short recording can introduce vocabulary, point at the policy, or refresh a rule people already practised. It is a decent handout. It is a poor substitute for the first time a team has to put a real file through a tool under a facilitator who will challenge them.
Use recorded content after a live cohort, not instead of one. People who have already practised a check can replay the standard. People who have never practised will click next until the bar fills.
Why Monday does not move
Behaviour at work changes when three things happen in the same week.
The work is in the room. Not a vendor demo file. A redacted version of the team's own notes, tickets, or drafts. Embedding AI in Daily Workflows is built that way: the unit of practice is a job people already do.
Someone can interrupt. A facilitator sees the bad paste, the invented citation, the skipped check. Peers hear the correction. That social proof is what an LMS cannot manufacture.
The manager is in the loop. If the line manager still rewards speed over a checked output, the module was theatre. Leading an AI-Ready Team is the session for that job. Champions programmes fail for the same reason: the person with the calendar and the one-to-ones was never trained. See Managers, not champions, after you buy the tools.
An LMS dashboard will not tell you those three things happened. It will tell you who clicked.
What good looks like
Commission live cohorts for the roles that already use the tools on customer or internal work. Keep the group small enough that a facilitator can watch the work. Take a register. Set an assessed task that looks like Monday, not like a multiple-choice recap of the slides.
Then, if you must, put a short recording in the LMS as the reminder: the checking standard, the policy line, the escalation. Do not invert that order.
If you are writing the brief, How to commission workforce AI training is the buying note. This one is the format note: live first when you need behaviour.
A good live design also produces evidence you can file: attendance, the task, the grade or observed outcome. That is what you will need if someone later asks whether the training stuck. A completion tick is not that evidence.
What to insist on in the brief
- Live facilitation, named, with a ratio that allows interruption.
- Practice on your work, or a close analogue, not only a public chatbot.
- Managers in a separate or overlapping session, not a forwarded link.
- A register and an assessed artefact the organisation keeps.
- Recorded modules only as follow-up, with that written in the proposal.
If a vendor's only offer is a platform seat and a content library, you are buying a library. Say so in the business case. Do not write "capability programme" on the same line.
Experrt runs live, in-house cohorts. We will not pretend a recording changed a shift. We will run the room, take the register, and leave you the pack.
How to start
Pick one team that already has licences and a visible quality problem: rework, unused seats, or outputs nobody will sign. Book a live cohort for that team and a manager session for the person they report to. Put any LMS module after those dates, as the reminder. If the only slot you can get is a recording, say so in the business case and do not call it a behaviour change.
Contact Experrt with the team, the tool, and what "different on Monday" would look like in one sentence. That is the brief.
Book a conversation
Experrt runs live, in-house cohorts. If this briefing matches a gap you already have, talk to us about scope, dates and the record the programme should produce.
Contact Experrt