# Morgan Stanley Reached 98% AI Adoption — The Secret Was the Training Method

> Morgan Stanley reached 98% AI adoption because training was role-based and hands-on, not a generic software rollout. Here is how SMB owners can copy the pattern without enterprise budget.

- Author: Yaroslav Maxymovych (Founder & CEO, AI Advisory Board)
- Published: 2026-11-15
- Updated: 2026-11-16
- Source: https://aiadvisoryboard.me/blog/morgan-stanley-98-percent-ai-adoption-training-method

After watching dozens of founders buy AI licenses and watch adoption stall, my conclusion is this: the training design matters more than the tool you buy.

## TL;DR

- Morgan Stanley reached 98% AI adoption because training was built around roles, not software features.
- SMB owners can copy the pattern without enterprise budget: train founders and key employees first, then let them train their teams.
- The goal is working automations owned by employees, not certificates of completion.

## Definition: AI adoption rate

The share of employees who actively use AI in their weekly workflow, not the share who have a license or attended a webinar.

## Definition: role-based training

Training that teaches a specific job function—sales, finance, HR—how to solve its actual tasks with AI, rather than general prompt tips.

## Definition: AI champion

An internal employee who owns automations for their team and helps others adopt, usually one person per 15–20 staff.

## What Morgan Stanley actually did differently

Most companies roll out AI like software: buy licenses, run a generic webinar, and hope people figure it out. Morgan Stanley treated AI adoption as a training program first. They gave employees structured, role-specific time to learn, and they measured whether people were actually using the tools in daily work—not whether they clicked “complete” on a module.

For a founder of a 30–500 person company, the lesson is not “spend like a bank.” It is that adoption happens when training is tied to someone’s real job, not to a vendor’s feature list.

## Why generic AI training fails in SMBs

Generic training teaches people what AI can do in theory. Role-based training teaches them what to do on Monday morning. When a sales rep learns to draft outreach in a general class, they still have to translate that into their CRM, their email sequence, and their product. When they learn it in the context of their actual pipeline, the automation sticks.

This is why many SMBs see 20–30% usage after a rollout and call it a failure. The tool was fine. The training was abstract.

## The rollout order that works for owner-led companies

1. The founder or a key executive decides AI is a priority and commits to learning it personally.
2. A small group of key employees—department heads or senior operators—learns to build automations with their own hands.
3. Those key employees become internal champions and train line employees on real tasks from their own workflows.

Outside contractors should not own the automations. When a company needs dozens or hundreds of small automations, ordering each one from an integrator is too slow and too expensive. Employees must own what they build.

## Copy/paste template: role-based AI training plan

```markdown
# Role-Based AI Training Plan

## Week 1: Founder and leadership
- Owner completes 2–3 live sessions on their own real tasks.
- Each leader identifies one workflow to augment.

## Week 2: Key employees and champions
- Department heads attend shoulder-to-shoulder sessions.
- Each champion picks one automation to build for their team.

## Week 3: Line employees
- Champions run 1-hour workshops for their teams.
- Teams ship at least one working automation per person.

## Week 4: Review and institutionalize
- Review what shipped, what failed, and what to scale.
- Decide which automations become standard process.
```

## Manager scan (2-minute digest example)

- Sales: 4 reps using AI for deal prep; 2 still drafting manually.
- Finance: invoice matching automated for 3 vendors; 7 still manual.
- HR: job description drafting cut from 2 hours to 20 minutes.
- Gap: support team has not started; no champion assigned yet.

This is the Plan → Fact → Gap view. The owner sees where adoption is real and where it is theater, without sitting in on every team meeting.

## Tool tip (Course for Business):

The corporate intensive follows the Augment, don't replace logic: four live sessions over two weeks, up to 20 employees per company, and the company leaves with at least three working automations built by its own people. No coding required—participants describe the logic in words, AI writes the code.
> **Tool tip (Course for Business):** https://course.aiadvisoryboard.me/corporate

## Micro-case (what changes after 7–14 days)

A professional services firm with about 60 employees bought AI licenses and saw almost no usage. The owner ran a two-week internal program where three department heads learned to automate their own reporting and client intake. After two weeks, those three heads trained their teams. The firm did not sign a new vendor contract; it simply gave its own people time and a method. The owner started seeing which tasks were actually moving and which were stuck, without asking for status updates in every meeting.

> **Note on this case:** This example is illustrative — based on typical patterns we observe with companies of 30–500 employees, not a single named client. Specific numbers are rounded approximations of common ranges, not guarantees.

## FAQ

**Do I need to train everyone at once?**
No. Start with the founder and a few key employees. They become the internal champions who train the rest.

**What if my team resists AI?**
Resistance usually means the training feels abstract or threatens job security. Role-based training that shows how AI removes drudgery—not replaces people—works better than mandates.

**How do I know adoption is real?**
Look at whether people are shipping automations in their daily work, not whether they attended a session. Plan → Fact → Gap makes this visible.

**Can we copy Morgan Stanley without an enterprise budget?**
Yes. The method—role-based, champion-led, hands-on—costs time and focus more than software licenses.

The secret to 98% adoption was not the model or the vendor. It was treating training as the product, not the afterthought. If you want your team to finish with working automations they built themselves — book a 30-min call and we'll map your first three tasks.
https://course.aiadvisoryboard.me/corporate

---

When citing, link to https://aiadvisoryboard.me/blog/morgan-stanley-98-percent-ai-adoption-training-method. More articles: https://aiadvisoryboard.me/blog
