Research
The Research Behind Every Course We Teach
Every AI-literacy course we teach is built on a working research base, tested against real rooms of real employees, and revised on a documented schedule — not written once and left alone. This page shows the work, with the numbers attached.
Three commitments behind the curriculum
What "we research this" actually means in practice, before any of the numbers below.
We research the curriculum
Every instrument we build traces to a working source list — worker-voice research, test-fairness literature, item-quality psychometrics, field precedents and labor-market evidence — each checked against the publishing body's own page, not a secondary summary.
We're grounded in both research and practice
Our skill domains are identified through external labor-market research as well as our own functional experience actually building and operating AI systems — how they work, where privacy and data-security risk actually shows up on the job. AI Catalyst Institute is Ohio registered school #2332, with six State of Ohio Registered Programs in applied AI skills.
We keep revising
Course material is continuously updated as new industry research and our own classroom data come in — it isn't written once and left alone.
Our own data, room by room
Every workshop, keynote and training session we run collects live poll data. Here's the aggregate, generalized — no organization identified.
59.7% of respondents already call themselves a regular AI user or beyond — before we ever teach them anything (n=481, aggregated across every engagement that asked this question)
*Reach, not unique headcount — each engagement's single largest per-question respondent count, summed. A person who attended two of our sessions is counted in both, not once. We're not saying "532 people trained."
How our learners move
Real first-session/last-session polls on the identical question, across our strongest cohorts.
35%→79%
of respondents report active AI use or high confidence (n=46→47, pooled across 3 cohorts)
Self-reported, first vs. final session.
24.1%→74.1%
Confidence using AI for a work task, opening poll vs. closing poll, same one-day training event (n=29→27)
an educational-partner session series (n=11→12)
Clear, large shift: every respondent in the later session self-identifies as an active AI user or beyond, up from about two-thirds at the start.
a recurring corporate training client (n=6→8)
Roughly doubles the share of active AI users.
a one-day conference / foundation-day training event (n=29→27)
The single strongest pre/post result in this set: confidence roughly triples in top-2-box share within one training day, on the exact same question asked twice.
Us vs. industry, side by side
Where we already have a real, sourced comparison — publisher, report and year on every figure.
Workforce upskilling coverage vs. our single-session lift
Industry — Boston Consulting Group
>50% vs 20%
Our rooms
35% → 79% (pooled, first vs. final session)
BCG frames workforce upskilling coverage as a hard ceiling most firms don't clear even at their best (roughly half their people, at best). Our own pooled pre/post shows individual cohorts moving past that ceiling's upper bound within a single short engagement - a complementary, not competing, framing: the industry number describes the scale of the problem, ours describes the size of the lift a session can produce.
AI training-access gap vs. our own room composition
Industry — The Conference Board
33.3%
Our rooms
40.3%
The Conference Board figure says roughly two-thirds of workers who already use AI never got formal employer training for it. Our own baseline shows a comparable-sized share of our room walks in not yet using AI regularly at all - i.e., our audience sits squarely inside the exact training gap industry data describes, and our training shows a measurable same-day close on it.
Frontline AI friction vs. our frontline same-day confidence gain
Industry — McKinsey (QuantumBlack)
47% vs 31%
Our rooms
24.1% to 74.1% (top-2-box confidence, same day)
McKinsey documents that AI-related strain concentrates on frontline/midlevel workers, not executives. Our sharpest same-day confidence gain comes from exactly that kind of frontline/general-employee session - a direct, complementary pairing: industry names where the pain is, our data shows a same-day move on it for that same population.
What different learners get out of it
What each group is built to get out of a session, with real quotes from our own poll open-text answers.
Executives
Built to give executives a clearer, faster basis for decisions — not another dashboard to interpret.
“I'm a business solutions professional. Without understanding these tools, I cannot provide clear guidance for the employers I serve.”
Advanced users
Built to reduce internal friction: when more people across an organization share real AI literacy, less gets re-explained and re-litigated.
“automating ticket resolution for repetitive tasks using client environments as custom variables.”
“Working with the team to verify Agents we built and share out”
“Standardization across the company with how we operate”
Beginners
Built to be approachable: no prior experience assumed, in a low-pressure room.
“I meant I like the class”
“Helping others better understand technology and feel more comfortable using it”
“Comfort with what I can do and what I can't do”
How that compares to industry research
Our own rooms, next to what the established research firms are finding. Every figure below is independently verified against the publisher's own page or report — publisher, report and year cited for each. (15 of 20 candidate figures cleared that bar; the rest are held back until independently verified.)
| Figure | What it measures | Source |
|---|---|---|
| 80% / 50% | Share of survey respondents reporting AI improved their individual productivity (80%); share reporting AI helps them make better decisions (50%) | McKinsey (QuantumBlack) The State of AI in 2026: On the Road to ROI, 2026 |
| 89% / 44% (up from 38%) | Share of organizations using AI in at least one business function (89%); share reporting AI scaling across the enterprise (44%, vs 38% a year earlier) | McKinsey (QuantumBlack) The State of AI in 2026: On the Road to ROI, 2026 |
| 47% vs 31% | Share of midlevel managers and individual contributors reporting at least one AI-related negative effect at work (47%), vs. executives and senior managers (31%) | McKinsey (QuantumBlack) The State of AI in 2026: On the Road to ROI, 2026 |
| 37% | Share of respondents whose organizations attribute at least some EBIT impact to AI use, essentially unchanged year over year | McKinsey (QuantumBlack) The State of AI in 2026: On the Road to ROI, 2026 |
| 6% | Share of survey respondents classified as 'AI high performers' (organizations attributing >=5% EBIT impact to AI and describing its value as 'significant'), unchanged from 2025 | McKinsey (QuantumBlack) The State of AI in 2026: On the Road to ROI, 2026 |
| 70% / 20% / 10% | Share of AI transformation value BCG attributes to the people component (70%) vs. the technology required to implement algorithms (20%) vs. the algorithms themselves (10%) | Boston Consulting Group AI Transformation Is a Workforce Transformation, 2026 |
| >50% vs 20% | Share of employees 'future-built' companies plan to upskill on AI (>50%) vs. companies BCG classifies as laggards (20%) | Boston Consulting Group AI Transformation Is a Workforce Transformation, 2026 |
| 85% vs 78%; $45,000 vs $30,000 | Full-time employment rate for adults with a certificate/certification vs. no credential (85% vs 78%); median annual income for the same comparison ($45,000 vs $30,000) | Strada Education Foundation (with Lumina Foundation and Gallup) Certified Value: When Do Adults Without Degrees Benefit from Earning Certificates and Certifications?, 2019pre-2024 context |
| 10% / 45% / 8% | Frontline-worker goal-achievement gain after training (10%); spillover effects on managers as a share of the program's total benefits (45%); productivity gain for supervisors working closest with trained workers (8%) | Harvard Business School (Working Knowledge) Why Employee Training Pays Off for Companies Twice, 2026 |
| No positive impact at 30 months; underemployed 4+ years post-layoff | WIA (Workforce Investment Act) Adult and Dislocated Worker training services' impact on earnings/employment 30 months after enrollment (none found, per national RCT); TAA (Trade Adjustment Assistance) participants' employment/earnings relative to non-TAA dislocated workers, including 4+ years after job loss | Brookings Institution AI Labor Displacement and the Limits of Worker Retraining, 2025 |
| +30% at 6 yrs; +$8,000/yr; +$38,000 cumulative; $2.46 per $1 | Year Up United treatment-group earnings vs. control group 6 years post-graduation (+30%); average annual earnings gain sustained through year 7 (+$8,000/yr); cumulative earnings gain over the study period (+$38,000); societal benefit per dollar invested ($2.46) | Year Up United PACE (Pathways for Advancing Careers and Education) evaluation, third report in series, 2022pre-2024 context |
| +32% (St. Nicks Alliance); no Year-10 impact (other 3 sites, incl. Per Scholas) | 10-year earnings outcomes across the four WorkAdvance sector-focused training program sites (Per Scholas, St. Nicks Alliance, Towards Employment, Madison Strategies Group) | MDRC Effects of Sector-Focused Training After 10 Years: Findings from the WorkAdvance Evaluation, 2025 |
| 88% / 70% | Share of surveyed organizations using AI in at least one capacity (88%); share using generative AI in at least one business function (70%) | Stanford HAI (Institute for Human-Centered AI) The 2026 AI Index Report, Economy chapter, 2026 |
| 55.1% / 33.3% | Share of workers who use generative AI or AI agents daily or weekly (55.1%); share who received organization-provided AI training in the past six months (33.3%) | The Conference Board AI-skilling research (press release), based on interviews with 35 enterprise leaders and a global survey of nearly 1,300 workers, 2026 |
| 92%; ~1/3 | Share of jobs analyzed that require digital skills (92%); share of workers who lack the foundational digital skills needed to enter and thrive in today's jobs (about one-third) | National Skills Coalition Closing the Digital Skill Divide, 2023pre-2024 context |
Figures marked pre-2024 context are real, useful comparators for the general value of training and credentials — they predate the current AI-adoption wave and are shown as historical baseline, not current AI-specific data.
Planning AI training for your Ohio team?
Our courses are pending TechCred approval. Contact us to learn more about training options.