Research & Evidence

The Evidence

Documented case studies, peer-reviewed research, & institutional reports from HBR, McKinsey, Stanford, Harvard, the World Economic Forum, & more. For L&D professionals who'd rather see the data than take someone's word for it.

People celebrating success — hand-drawn illustration

The Big Picture

Six Patterns Worth Watching

Hyper-Personalization at Scale

AI algorithms create individualized learning paths based on job roles, existing skills, performance metrics, & career aspirations — moving past static, one-size-fits-all programs.

Simulation-Based Practice

Tools like Yoodli, Mursion, and Strivr enable realistic practice scenarios with AI feedback — from sales conversations to difficult management situations — available 24/7.

Accelerated Content Creation

Generative AI is cutting content development time by 50-80%, letting L&D teams rapidly generate course outlines, scripts, assessments, & interactive simulations.

The Judgment Premium

As AI handles routine cognitive tasks, the premium on human judgment, critical thinking, & ethical reasoning goes up. Training must develop these meta-skills — the ones AI can't replicate.

AI Coaching at Scale

ATD reports 96% of users say AI coaches provide custom feedback tailored to their goals, and 89% say it results in specific, useful next steps. AI can even detect systemic organizational issues across multiple coachees.

The 'Being Side' of Leadership

Dan Hodgkins argues in ATD that the 'being side' — emotional intelligence, self-awareness, presence — is now the true differentiator in leadership development, not the 'doing side' that AI can replicate.

Case Studies

Who's Actually Doing This Well

Documented examples from organizations that have integrated AI into their training programs — with real tools, real numbers, & real lessons learned.

Technology

IBM

The Challenge

Training a vast, diverse global workforce when skill requirements shift faster than any curriculum committee can keep up with.

The Solution

Put Watson AI at the core of corporate training. Watson analyzes employee data — job roles, past performance, completed training, & career goals — to construct individualized learning paths & perform skills gap analyses that would take human teams months to replicate.

Tools used: IBM Watson (proprietary)

The Results

Substantial reduction in training time
Marked increase in employee satisfaction
Higher course completion rates
More precise skills gap identification

Key Lesson

Personalization at scale works when AI handles the data analysis & L&D teams focus on what they're actually good at: content quality & strategic alignment.

Pause & Reflect

IBM used AI to analyze job roles, performance, & career goals. What data does your organization already have about employees that nobody is using for learning design?

Experiment to Try

Pick five employees with different roles. Sketch what an AI-personalized learning path would look like for each. Where do the paths diverge? That divergence is where personalization actually matters.

Source: eLearning Industry
Retail

Walmart

The Challenge

Training a massive retail workforce on high-pressure scenarios — Black Friday surges, difficult customer interactions — in a way that's realistic, scalable, & actually measurable.

The Solution

Partnered with STRIVR to deploy AI-powered VR training across academies & stores. The AI tracks learner performance inside VR simulations — gaze, reaction time, decision-making — & delivers personalized, real-time feedback.

Tools used: STRIVR (VR + AI)

The Results

15% improvement in employee performance
95% reduction in training time for certain tasks
More confident employees in high-pressure situations
Scaled across 1.5M+ employees

Key Lesson

Immersive, AI-powered simulations compress months of on-the-job learning into hours. Especially for experiential skills that no classroom lecture has ever successfully taught.

Pause & Reflect

Walmart trained for Black Friday surges in VR. What's your organization's 'Black Friday' — the high-pressure scenario everyone dreads but nobody practices for?

Experiment to Try

Identify one high-stakes scenario your team faces. Script a 10-minute roleplay version of it. Run it with three people. Then ask: what would change if AI could give them real-time feedback during the practice?

Source: eLearning Industry
Logistics & Technology

Amazon

The Challenge

Training warehouse staff to work safely & efficiently alongside AI-powered robots in highly automated fulfillment centers.

The Solution

Built AI-enhanced training modules that adapt to individual learning pace. When a worker struggles with a particular task, the AI steps in — real-time guidance & visual aids. The system tracks task completion time, accuracy rates, & movement patterns.

Tools used: Custom AI training platform

The Results

75% increase in employee engagement
40% improvement in task fulfillment time
Reduced safety incidents
Adaptive difficulty based on individual progress

Key Lesson

When humans & AI work side-by-side, the training itself should model that collaboration. Adaptive, responsive, & personalized to each worker's actual needs.

Pause & Reflect

Amazon's training adapts in real-time based on where workers struggle. Does your training do that — or does everyone get the same content regardless of what they already know?

Experiment to Try

Take your most common training module. Identify the three points where people usually get stuck. Design a branching path: if someone nails the concept, skip ahead. If they struggle, offer a different explanation. You've just built the logic for adaptive learning.

Source: eLearning Industry
Professional Services

Deloitte

The Challenge

Systematically upskilling a massive consulting workforce in AI & emerging technologies to maintain competitive advantage — & actually serve clients effectively.

The Solution

Created the Deloitte AI Academy, providing hands-on, experiential training organized around real-world use cases. The curriculum ensures training is practical and immediately applicable. Deloitte now offers its AI Academy as a service to clients.

Tools used: Deloitte AI Academy (proprietary)

The Results

Workforce-wide AI literacy improvement
New revenue stream (AI Academy as a service)
Practical, use-case-driven curriculum
Hands-on experiential learning model

Key Lesson

The most effective AI training is organized around real use cases, not abstract concepts. Build something good enough for your own people & it might become a product. Deloitte noticed.

Pause & Reflect

Deloitte built their AI Academy around real use cases, not abstract concepts. If you had to pick three real tasks your people do every week that AI could transform — what are they?

Experiment to Try

Take one of those three tasks. Spend 30 minutes doing it the old way, then 30 minutes with AI assistance. Document the difference — not just in time saved, but in quality of output. That's your first use case.

Source: Deloitte
Consumer Goods

Unilever

The Challenge

Streamlining onboarding for new hires across a global organization — cutting time-to-productivity while keeping the welcome warm & human.

The Solution

Developed Unabot, an AI-powered chatbot built on Microsoft's Bot Framework with NLP capabilities. Unabot serves as a virtual assistant for new hires — answering questions about ID badges, benefits, company values, & culture.

Tools used: Microsoft Bot Framework (Unabot)

The Results

High adoption rates among new hires
Strong perception of usefulness
Reduced burden on HR teams
24/7 availability for new employee questions

Key Lesson

AI chatbots handle the routine questions that consume HR & L&D time. The key: making them feel helpful & human — not like a deflection mechanism.

Pause & Reflect

Unilever's chatbot handles routine onboarding questions. What are the ten questions your HR team answers most often? How many of those could a well-designed chatbot handle without making people feel dismissed?

Experiment to Try

Collect the 20 most common questions new hires ask in their first week. Feed them to ChatGPT with your company context. Rate the answers. The ones it nails? Those are your chatbot candidates. The ones it botches? Those stay human.

Source: eLearning Industry
Talent Advisory

Korn Ferry + Yoodli

The Challenge

Scaling high-quality communication coaching across a global client base without proportionally scaling the number of human coaches. The math doesn't work otherwise.

The Solution

Partnered with Yoodli to provide AI-powered communication coaching. The platform enables realistic roleplay practice for sales conversations, presentations, & interviews — with real-time feedback on pacing, filler words, confidence, & clarity.

Tools used: Yoodli (AI communication coach)

The Results

Scalable coaching without proportional headcount
Real-time feedback on speaking skills
Practice available 24/7, not just during coaching sessions
Bryan Ackermann called it 'the perfect augmentation to human coaching'

Key Lesson

AI coaching works best as augmentation, not replacement. AI practice + human guidance produces better outcomes than either alone. Bryan Ackermann called it 'the perfect augmentation to human coaching.' The data backs him up.

Pause & Reflect

Korn Ferry uses AI for practice reps between human coaching sessions. What skill in your organization would benefit most from unlimited, judgment-free practice?

Experiment to Try

Record yourself giving a 3-minute presentation using Yoodli's free tier. Review the AI feedback on filler words, pacing, & confidence. Then give the same presentation again. The improvement in one cycle will tell you everything about the value of AI-assisted practice.

Source: Yoodli
Financial Services

JPMorgan Chase

The Challenge

Reskilling 300,000+ employees for an AI-transformed financial services landscape — while maintaining regulatory compliance & operational continuity.

The Solution

Committed to reskilling its entire workforce with AI-focused training. The bank's approach emphasizes practical application — employees learn AI tools in the context of their actual job functions, from risk analysis to customer service.

Tools used: Internal AI training platform + external partnerships

The Results

Workforce-wide AI literacy initiative
AI integrated into daily workflows
Practical, role-specific training paths
Leadership commitment from CEO Jamie Dimon

Key Lesson

Top-down commitment matters. When the CEO publicly champions AI reskilling, it signals to the entire organization that this isn't optional. Jamie Dimon didn't send a memo — he made it a strategic priority.

Pause & Reflect

Jamie Dimon made AI reskilling a visible strategic priority. Who in your organization's leadership is visibly using AI? If the answer is 'nobody' — that's your adoption problem in one word.

Experiment to Try

Draft a one-paragraph 'AI commitment statement' from your CEO's perspective. What would it say? Now ask yourself: would your actual CEO sign it? The gap between those two documents is your change management challenge.

Source: Reuters / JPMorgan Annual Report

Academic & Institutional Research

What the Research Actually Says

Peer-reviewed papers & institutional reports — not blog summaries or vendor marketing. Each entry includes the key finding & its practical implication for your work.

Learning Science · 2025

The Impact of AI on Workplace Learning: A Systematic Review

Barcaui, A. & Werner, A. — International Journal of Educational Technology

Key Finding

Students who used ChatGPT as a study aid scored 11 percentage points lower on subsequent tests compared to those who studied without AI assistance.

So What? (Implication for You)

AI can create dependency that undermines deep learning. The fix isn't banning AI — it's designing for 'desirable difficulties,' the productive struggle that builds genuine expertise. Design matters more than access.

Workforce Transformation · 2025

Generative AI and the Future of Work

McKinsey Global Institute — McKinsey & Company

Key Finding

AI could automate 60-70% of current work activities. The long-term economic opportunity is estimated at $4.4 trillion in added productivity annually.

So What? (Implication for You)

The scale of workforce transformation requires L&D to move from periodic training events to continuous learning ecosystems. Only 1% of leaders consider their organizations 'mature' on AI deployment. One percent.

Human-AI Collaboration · 2024

AI Won't Replace Humans — But Humans With AI Will Replace Humans Without AI

Karim Lakhani — Harvard Business School

Key Finding

In controlled experiments, consultants using AI outperformed those without by 40% on quality metrics and completed tasks 25% faster.

So What? (Implication for You)

The competitive advantage isn't AI itself — it's the human ability to collaborate with AI effectively. Training should focus on human-AI collaboration skills, not tool proficiency alone.

Psychological Safety · 2026

The Psychological Safety Imperative in AI-Augmented Teams

Edmondson, A. C. & Seth, J. — Harvard Business Review

Key Finding

AI creates 'trust ambiguity' by providing confident but potentially incorrect information. Teams with high psychological safety are 3x more likely to effectively challenge AI outputs.

So What? (Implication for You)

Building psychological safety is now more critical than ever. Facilitators need to create environments where people feel safe questioning both humans & machines.

Workforce Transformation · 2025

The State of AI in 2025

McKinsey & Company — McKinsey Global Survey

Key Finding

92% of companies plan to increase AI investments over the next three years. Yet 74% report struggling to achieve and scale value from AI initiatives.

So What? (Implication for You)

The gap between AI investment & AI value is a people problem, not a technology problem. L&D professionals are uniquely positioned to close it — through effective training & change management.

Human-AI Collaboration · 2023

Navigating the Jagged Technological Frontier

Dell'Acqua, F. et al. — Harvard Business School / BCG

Key Finding

BCG consultants using GPT-4 improved performance by 40% on tasks inside AI's capability frontier, but performed 23% worse on tasks outside it — falling into 'automation complacency.'

So What? (Implication for You)

AI has a 'jagged frontier' — brilliant at some tasks, terrible at others, & the boundary isn't obvious. Training needs to help people develop judgment about when to trust AI & when to override it.

Skills & Competencies · 2025

AI and the Future of Skills

World Economic Forum — World Economic Forum

Key Finding

By 2030, 39% of core workplace skills will change. Critical thinking, creative problem-solving, and AI literacy are the fastest-growing skill demands.

So What? (Implication for You)

The skills that matter most in an AI world are deeply human: judgment, creativity, empathy, & the ability to ask the right questions. Those are facilitation skills.

Learning Science · 2024

Artificial Intelligence in Corporate Training: A Comprehensive Review

Multiple authors — Journal of Workplace Learning

Key Finding

AI-powered adaptive learning systems show 30-50% improvement in knowledge retention compared to traditional e-learning, but only when combined with human facilitation and social learning.

So What? (Implication for You)

AI alone doesn't improve learning outcomes. AI personalization + human facilitation + social learning is what drives results. Take any one element away & the effect collapses.

Organizational Change · 2025

The AI-First Playbook

Board of Innovation — Board of Innovation

Key Finding

Despite $30-40B in GenAI investment, 95% of organizations are getting zero measurable return. The primary barrier is not technology but organizational readiness and change management.

So What? (Implication for You)

For every dollar in AI technology, invest two in people & change management. The organizations that win aren't the ones with the best tools — they're the ones that help their people adapt.

Learning Science · 2024

Personalized Learning at Scale: What the Evidence Says

Pane, J. F. et al. — RAND Corporation

Key Finding

Personalized learning approaches show positive but modest effects on student achievement. The largest gains come when personalization is combined with human mentoring and peer collaboration.

So What? (Implication for You)

Personalization is powerful but not sufficient on its own. The human elements — mentoring, coaching, peer learning — remain essential multipliers. Technology without human connection? Expensive content delivery.

From the Knowmium Workshop

How We're Actually Using AI

How we actually integrate AI into workshop design, facilitation, & content creation at Knowmium. Real tools, real workflows, real results.

Expanding Soon

AI-Powered Simulation Practice with Yoodli

How We Use AI Roleplay to Build Real Communication Skills

Yoodli is phenomenal for simulation-based practice. We use it to give workshop participants realistic, repeatable practice environments for presentations, difficult conversations, sales pitches, & interview prep — with instant AI feedback on pacing, filler words, confidence, & clarity. People can practice 50 times without burning out a human coach. The AI catches patterns that humans miss. This section walks through how we integrate Yoodli into our workshop design, the specific exercises we run, & the results we're seeing.

Expanding Soon

Microlearning Creation at Speed

From Idea to Deployable Content in Hours, Not Weeks

Traditional microlearning creation takes 40-100 hours per finished hour of content. With AI, we've compressed that to a fraction. We use generative AI to draft scripts, create scenario branches, generate assessment questions, & produce visual assets — then layer in human expertise for quality control, brand voice, & pedagogical rigor. This section details our exact workflow, the tools we use at each stage, & the quality assurance process that keeps AI-generated content teaching rather than just informing.

Expanding Soon

Deep Personalization in Practice

Making Every Learner's Path Genuinely Individual

Not 'choose your own adventure' branching — AI that analyzes a learner's role, existing skills, performance data, & career goals to construct genuinely individual learning paths in real time. This section shows how we implement deep personalization in our workshops & programs, the specific AI tools we use, & the measurable difference it makes versus one-size-fits-all approaches.

Expanding Soon

Spinning Up Purpose-Built AI Tools

Custom AI Applications for Specific Training Needs

The current AI landscape lets you spin up purpose-built tools for specific training needs — fast. Need a roleplay partner that understands your company's product line? Build one. A coach that gives feedback based on your competency framework? Build one. A knowledge base for your specific policies? Build one. This section walks through our process for identifying where purpose-built AI tools add the most value, how we build them (often in hours, not months), & how we fold them into broader programs.

"AI is poised to reinvent the $400 billion corporate training market. Our entire approach, philosophies, tech stack, and operating models for learning are out of date."

Josh Bersin, Global Industry Analyst