AI Powered Futures Final Report

Average attendance

77%

Across all three cohorts

Feedback responses

203

Learner responses across 8 modules

Usefulness: rated 4 or 5

78%

Of all program-wide responses

Knowledge check range

84%

To 96% across assessments run

Live sessions delivered

24

Across all three cohorts

Distinct content types

28

All custom built for CILAR

Self-paced lessons built

255+

Across three tailored course builds

Avg AI confidence growth

+2.0

+1.55 to +2.65 pts across cohorts

Five core frameworks taught progressively across all eight modules

01

DECIDE Check

02

AI Responsibility Gate

03

AI Decision-Making Matrix

04

AI Output Evaluation Lens

05

AI Responsibility Level Pyramid

About the program

AI Powered Futures is an eight-week blended learning program built to develop practical AI literacy for retail and customer-facing professionals. Developed by Wiz Learning for CILAR in partnership with Toronto Metropolitan University's G. Raymond Chang School of Continuing Education, the program runs across eight live sessions and a corresponding self-paced track on the Disco learning management system, with completion supporting eligibility for a TMU microcredential. Three pilot cohorts ran between March and June 2026, each across eight weeks and 40 hours of blended instruction.

Modules

8

Weeks

8

Hours

40

Cohorts

3

LMS

Disco

Cohort overview

Avg attendance

Cohort 1

Retail staff cohort · March 18 to May 6 · 24 enrolled

81%

Cohort 2

Team lead and manager cohort · April 8 to May 27 · 16 enrolled

84%

Cohort 3

Mixed cohort · April 28 to June 4 · 17 enrolled

65.6%

Figures shown are average attendance across all 8 weeks. All cohorts ran virtually after Cohort 1's in-person Week 1.

Engagement composite by cohort

Cohort 1 (4.0/5)

Cohort 2 (4.49/5)

Cohort 3 (3.75/5)

Strengths across the pilot

Frameworks transferred to real work in every cohort. Learners in all three cohorts reported applying the AI Decision-Making Matrix to real situations at work before the program concluded. This is the clearest evidence the content did its job.

Knowledge retention was strong wherever assessments were run. Knowledge checks landed in the 84 to 96% range across mid- and end-of-program assessments in Cohorts 2 and 3. Cohort 1's post-program quiz average rose from 80% at baseline to 92% at completion.

AI confidence grew meaningfully in every cohort where growth data was collected. Cohort 1 gained 1.9 points, Cohort 2 gained 2.65 points, and Cohort 3 gained 1.55 points on a 10-point confidence scale. All three cohorts described the change as significant in their end-of-program reflections.

Cohort 2 was the strongest pilot on every engagement measure. At 84% attendance, 4.49/5 engagement, and 89% of usefulness ratings at 4 or 5, Cohort 2 set the high-water mark for the program. The deliberate practice of grouping learners by career similarity in breakout rooms, which that cohort originated, is now a standing facilitation principle for all future cohorts.

The program proved transferable well beyond its original retail-only design. All three cohorts included substantial numbers of learners from corporate, banking, government, hospitality, and tech. Cohort 3 was the most industry-diverse pilot yet. Across all three groups, the applied decision-making framing landed regardless of sector.

The core message held from Week 1 through Week 8 in every cohort. Learners across all three groups consistently echoed the program's framing: human judgment belongs in every AI-assisted decision. The clearest expression of this came from a Cohort 3 learner in the program's final session.

Module 6 and Module 3 were standout sessions across every cohort. Module 6 (adapting AI skills across roles) was among the highest-rated sessions in all three cohorts. Module 3 (prompt engineering) was repeatedly named the most practically useful module by learners in their end-of-program reflections.

Early self-paced completion was the strongest of the program in every cohort. Modules 1 and 2 completion ranged from 60 to 86% depending on cohort. For Cohort 1, the largest concept familiarity gains landed in exactly the areas learners flagged as their weakest at the start: prompt engineering, data privacy, and bias and fairness.

Areas the program is actively addressing for future cohorts

Self-paced completion for Modules 7 and 8. Completion dropped to 27 to 43% in the final two modules across all three cohorts. This gap is the most consistent finding of the pilot and the one most directly tied to microcredential readiness. Earlier structured reminders, tied explicitly to assessment requirements, are already planned for future cohorts.

Final-week attendance. Attendance in the last two sessions was the lowest of the program in two of three cohorts, despite those sessions carrying the microcredential preparation lab and wrap-up. A communication strategy targeting advance notice for the final two sessions is being built into the facilitation standard for future cohorts.

Cohort 3 attendance and engagement. Cohort 3's 65.6% average attendance and 51% usefulness ratings of 4 or 5 stand apart from Cohorts 1 and 2. Knowledge check performance remained strong (84 to 96%), which points to an engagement and attendance gap rather than a content gap. Attendance monitoring and a check-in process for absent learners are being added for future cohorts.

Feedback collection in the second half of the program. Response rates fell sharply in later weeks across all cohorts, most severely in Cohort 3. Building the feedback form into the final five minutes of every session as a structured close, rather than a post-class ask, is a straightforward fix being implemented.

Module 4 format. Module 4 (responsible and ethical AI use) coincided with engagement dips in two of three cohorts. Content understanding remained strong wherever it was measured. The format and opening structure of the session are under review ahead of the next cohort.

Cohort composition variability. Each pilot included learners from a wide range of industries, roles, and experience levels. A more cohesive learner demographic within each cohort would allow for better tailoring of scenarios and pacing. Intake and cohort-sorting guidance is being built into the Program-in-a-Box handoff.

"Human is always the final responsible person, not AI. I still have to use my knowledge to judge the answer given by AI."

Cohort 3 · Module 8 · Program wrap-up

"I was unconfident and struggled to use my curiosity. Now I am much more confident and structured and I have made a lot of headway in the space."

Cameron J. · Cohort 2 · Post-program reflection

"I was comfortable with AI prior to this program. I am still comfortable with AI but now I am more aware and intentional. Now I use my human judgment to determine if it is necessary to use AI to help me complete a task."

Ephrata G. · Cohort 3 · Post-program reflection

"When I started this program, I already felt fairly comfortable using AI at work. However, I was often just plugging things in without fully understanding how to use AI effectively, thoughtfully, or responsibly. Now I feel more intentional and confident."

Melody L. · Cohort 1 · Post-program reflection

Cohort 1 avg attendance

81%

Retail staff cohort · 24 enrolled

Cohort 2 avg attendance

84%

Team leads and managers · 16 enrolled

Cohort 3 avg attendance

65.6%

Mixed cohort · 17 enrolled

Attendance across all three cohorts by week

Cohort 1

Cohort 2

Cohort 3

Cohort 2 Week 1 denominator is 15 (two learners withdrew after Week 2, bringing active enrolment to 14). Cohort 3 denominators: 17 in Week 1, 16 in Weeks 2 and 3, 15 from Week 4 onward.

Weekly engagement composite across all three cohorts (out of 5)

Cohort 1 (4.0 avg)

Cohort 2 (4.49 avg)

Cohort 3 (3.75 avg)

Composite = instructor rating 40% (engagement + comfort participating averaged) + student usefulness 35% + student understanding 25%. All three lines share the same formula and scale.

Cohort 1 · Usefulness

0

1

2

10

3

33

4

25

5

84% rated 4 or 5

Avg 4.19/5 · 69 responses

Cohort 2 · Usefulness

0

1

2

7

3

31

4

31

5

89% rated 4 or 5

Avg 4.31/5 · 70 responses

Cohort 3 · Usefulness

1

0

4

2

15

3

15

4

6

5

51% rated 4 or 5

Avg 3.49/5 · 41 responses · One 0/5 rating from a learner whose stated interests were outside program scope

Self-paced module completion across all three cohorts (%)

Cohort 1

Cohort 2

Cohort 3

The same declining pattern appears across all three cohorts. Modules 1 and 2 were consistently the strongest. Modules 7 and 8 were the lowest in every cohort, landing between 27 and 43%. The completion gap in the final modules is the most consistent finding of the pilot and is directly tied to microcredential readiness.

Cohort 1 · March 18 to May 6, 2026

Retail staff cohort

Applied AI for operational and service excellence

Enrolled24

Avg attendance81%

Engagement composite4.0/5

Usefulness 4 or 584%

Avg useful score4.19/5

Feedback responses69

Knowledge check: pre-program80% (4.8/6)

Knowledge check: mid-program94.1% (9.4/10)

Knowledge check: end-of-program92% (5.5/6)

AI confidence growth+1.9 pts

Self-paced M1/M280%/79%

Self-paced M7/M829%/29%

Microcredential9/10 passed

Lead instructorShawna M.

Co-instructorRyan P.

Cohort 2 · April 8 to May 27, 2026

Team lead and manager cohort

Leading responsible AI use in the workplace

Enrolled16

Avg attendance84%

Engagement composite4.49/5

Usefulness 4 or 589%

Avg useful score4.31/5

Feedback responses70

Knowledge check: pre-program85% (5.1/6)

Knowledge check: mid-program89% (8.9/10)

Knowledge check: end-of-program85% (13.6/16)

AI confidence growth+2.65 pts

Self-paced M1/M286%/86%

Self-paced M7/M843%/29%

Microcredential2/3 confirmed

Lead instructorMerveille M.

Co-instructorRyan P.

Cohort 3 · April 28 to June 4, 2026

Mixed cohort

Retail, tech, hospitality, consulting, and career transition roles

Enrolled17

Avg attendance65.6%

Engagement composite3.75/5

Usefulness 4 or 551%

Avg useful score3.49/5

Feedback responses41

Knowledge check: pre-program91.7% (5.5/6)

Knowledge check: mid-program96% (9.6/10)

Knowledge check: end-of-program84% (13.4/16)

AI confidence growth+1.55 pts

Self-paced M1/M260%/60%

Self-paced M7/M827%/27%

MicrocredentialResults pending

Lead instructorShawna M.

Co-instructorRyan P.

Cohort composition: who was in the room

Each cohort drew from a wider professional range than the original retail-only design anticipated. The program's applied, human-in-the-loop framing transferred well across all three groups regardless of industry background.

Cohort 1 backgrounds

Corporate and office roles · In-store retail · Warehousing · Banking · Nonprofits · Hospitality · Government

Cohort 2 backgrounds

Retail managers and team leads · Corporate office · Banking and financial services · Tech · Government · Nonprofit

Cohort 3 backgrounds

Tech · In-store retail · Hospitality · Consulting · Banking · Government · Logistics · Career transitioners

🎓

TMU microcredential · 100% pass rate on every graded submission

Every learner who submitted a graded assessment across Cohorts 1 and 2 passed. Eleven submissions, eleven confirmed passes. Cohort 3 results are pending. The TMU microcredential pathway includes two streams: CURV 700 Applied AI for Operational and Service Excellence (retail staff cohorts) and CURV 701 Leading Responsible AI Use in the Workplace (team lead and manager cohorts). Both require demonstrated applied AI competency, not participation alone.

Knowledge check performance across all three cohorts

Cohort 1

Cohort 2

Cohort 3

These three assessments differ in length and scope rather than testing the same material, so percentages are not directly comparable across phases. The end-of-program assessment is the most comprehensive, covering all eight modules and five frameworks. Completion rates for these optional assessments declined over the course of the program, consistent with the self-paced completion pattern shown elsewhere in this dashboard.

Cohort 1

Pre-program: 80% (4.8/6, n=24)
Mid-program: 94.1% (9.4/10, n=11)
End-of-program: 92% (5.5/6, n=6)

Cohort 2

Pre-program: 85% (5.1/6, n=12)
Mid-program: 89% (8.9/10, n=9)
End-of-program: 85% (13.6/16, n=7)

Cohort 3

Pre-program: 91.7% (5.5/6, n=11)
Mid-program: 96% (9.6/10, n=7)
End-of-program: 84% (13.4/16, n=5)

AI confidence growth (out of 10)

Before

C1 after

C2 after

C3 after

Growth reflects learners who completed both pre- and post-program surveys: 6 respondents for Cohort 1, 4 for Cohort 2, and 4 for Cohort 3. Cohort 3's baseline point uses the full 11-respondent baseline average, since individual baseline scores for those 4 post-program respondents were not isolated. Sample sizes are small; read these as directional trends rather than precise measurements.

Concept familiarity: where learners started and where they landed

Across all cohorts, the three concepts learners rated lowest at program start — prompt engineering, data privacy, and bias and fairness — consistently showed the largest gains by program end.

Prompt engineering

Start: low

Prompt engineering

End: high

Data privacy in AI

Start: low

Data privacy in AI

End: high

Bias and fairness

Start: low

Bias and fairness

End: higher

Based on Cohort 1 pre- and post-program familiarity data (0 to 4 scale). Cohort 2 post-program data reflects same pattern.

What learners said about their growth: across all three cohorts

"Human is always the final responsible person, not AI. I still have to use my knowledge to judge the answer given by AI."

Cohort 3 · Module 8 · Final session

"I was unconfident and struggled to use my curiosity. Now I am much more confident and structured and I have made a lot of headway in the space."

Cameron J. · Cohort 2 · Post-program reflection

"I was comfortable with AI prior to this program. I am still comfortable with AI but now I am more aware and intentional. Now I use my human judgment to determine if it is necessary to use AI to help me complete a task."

Ephrata G. · Cohort 3 · Post-program reflection

"When I started, I already felt fairly comfortable using AI at work because I had been using it regularly for over a year. However, I was often just plugging things in without fully understanding how to use AI effectively, thoughtfully, or responsibly. Now I feel more intentional and confident."

Melody L. · Cohort 1 · Post-program reflection

"I had thought AI would replace humans and take their jobs. For now, it seems like humans are still needed. Context, critical thinking, reasoning, and judgment are still required."

Samantha K. · Cohort 2 · Post-program reflection

"Making sure to revise and use my critical thinking skills and judgment, lead with empathy whenever I work with AI."

Caitlin M. · Cohort 3 · Post-program reflection

Skills and habits learners plan to take forward

AI Decision-Making Matrix DECIDE Check Prompt engineering Critical review of AI outputs Human in the loop AI Output Evaluation Lens Responsible AI in daily work Protecting sensitive data AI for documentation Scaffolding AI decision-making Reviewing before sharing AI Responsibility Gate

What learners want to learn more about

Prompt engineering depth AI tools by use case Durable skills alongside AI Bias identification AI in workflows and operations AI ethics and policy Agentic AI Legislative frameworks AI in e-commerce AI for business productivity AI in corporate settings Supply chain and inventory

What learners would tell someone starting the program

"Approach AI as a tool to support your thinking and not replace it."

Shauna E. · Cohort 1

"Don't worry if it feels unfamiliar at the beginning. Once you understand how to communicate with AI properly, it becomes a very practical and useful tool."

Neslihan M. · Cohort 2

"The program helps you understand how to navigate AI and use critical thinking skills."

Ephrata G. · Cohort 3

"You may have some knowledge about AI but this course expands your knowledge with ways to use AI practically and how to use it responsibly."

Tamara S. · Cohort 1

"Be prepared to be engaged in class activities and have an open mind."

Colin S. · Cohort 2

"It is so important to consider how AI will impact the future of the workplace, no matter what industry you're in."

Caitlin M. · Cohort 3

Full intellectual property ownership vests in CILAR

Every piece of content, every lesson, every framework application, every video, every assessment, and every deliverable in the program is owned outright by CILAR upon completion of the engagement. No pre-packaged or repurposed Wiz Learning content was used. Everything was designed and built specifically for this program.

■

Live session materials (per module, across 8 modules)

Module overviews (instructor and student versions) · Lesson plans with timed facilitation notes · Instructor presentation decks · Student presentation decks · Lesson overview sheets · Retail Scenario Labs (6 core sets, 26 scenario variations, instructor and student copies) · 8 additional lab sets for non-retail learners

■

Self-paced and digital course content

Three tailored Disco course builds (retail staff, team leads and managers, mixed cohort) · 255 lessons across all three builds · Disco interactive activities (one per module) · Four knowledge check quizzes built around applied reasoning, not recall

■

Assessments and feedback tools

Baseline and growth surveys · Eight module feedback forms · End-of-program assessment · Two Mock Microcredentials (retail staff version and team lead and manager version) · Post-program feedback form

■

Reporting and analytics

Three cohort dashboards, each with ten interactive reporting views · Three cohort summary reports · Consolidated program dashboard (this dashboard)

■

Multimedia

Four concept breakdown audio clips (podcast-style, paired with workplace examples) · Eight module audio recaps · Eight module overview videos · Three animated explainer videos for cohort courses · Ten animated explainer videos for the On Demand course

■

AI Powered Futures On Demand

Fully self-paced standalone version of the complete program · 113 lessons across all 8 modules · Seven knowledge checks · Ten animated explainer videos · Dedicated start- and end-of-program assessments · Own mock microcredential

■

Reference, support, and instructor documents

Student Review Lessons (module wrap-ups) · Course Review Document (student copy) · Key Concept Guide and Glossary · Three instructor-facing course overview documents · Instructional Approach and Pedagogy Guide · Module Learning Goals Packet · Evaluation Model and Participant Evaluation Framework

■

Train-the-Trainer and sustainability

Five-session Train-the-Trainer series with full verbatim facilitator scripts · AI Reader Packet · Facilitator Readiness Checklist · Program-in-a-Box: reusable DISCO template for future cohorts

24

Live sessions delivered across three cohorts

57

Learners enrolled across the full pilot

43

Meetings supporting program delivery and advisory work

5

Core frameworks taught progressively across 8 modules

8

Additional scenario lab sets built for non-retail learners

2

Mock Microcredential versions (retail staff and managers)

3

Tailored self-paced course builds for different audiences

203

Learner feedback responses collected across the pilot

The pilot built more than a program

Three cohorts, 57 learners, 24 live sessions, and 203 pieces of learner feedback have produced something larger than any single cohort outcome. AI Powered Futures now has a full content library, a proven delivery model, a sustainability infrastructure, and an evidence base that positions the program for independent delivery, broader reach, and continued iteration. Everything described here is already built. The question now is how to deploy it.

Run it independently

📷

Train-the-Trainer series

A five-session series built specifically to prepare CILAR's own facilitators to deliver the program without external support. Sessions cover the program overview and AI foundations, a full curriculum deep dive across all eight modules, facilitation standards and learner support, and adapting delivery across different contexts. Each session includes verbatim facilitator scripts, discussion questions, and observations drawn from all three pilot cohorts. Three sessions are complete and two are in progress.

📦

Program-in-a-Box

A reusable, templated version of the full live course build in Disco. Future cohorts can launch without rebuilding the program from scratch. CILAR's team can update content, adjust for new audiences, and run the program on its own timeline using the existing infrastructure.

☑

Facilitator Readiness Checklist and AI Reader Packet

A standalone checklist that helps new facilitators confirm they understand the material, frameworks, and delivery approach well enough to run the program independently. Paired with an AI Reader Packet — a structured document covering program structure, outcomes, pedagogy, and facilitation guidance, formatted to be kept current and maintained with AI assistance going forward.

Reach more learners

🌐

AI Powered Futures On Demand

A fully self-paced standalone version of the complete program, already built and ready to deploy for learners outside the live cohort structure. Includes 113 lessons across all eight modules, seven knowledge checks, ten animated explainer videos, dedicated start- and end-of-program assessments, and its own mock microcredential. This course runs without an instructor and can serve learners on their own schedule, opening the program to audiences that live cohorts cannot easily reach.

🏫

Three tailored builds ready for different audiences

The self-paced content exists in three distinct versions: one built for retail staff, one for managers and team leads, and one for mixed cohorts. Each build adapts scenarios, framing, and examples to the audience it serves. These builds are already complete and can be deployed for future cohorts without additional development work.

⇄

Content proven across industries

All three pilots demonstrated that the program's applied decision-making and responsible AI framing transfers well beyond retail. Learners from banking, tech, government, hospitality, consulting, logistics, and corporate office roles engaged meaningfully with the content and reported direct workplace application. The scenario labs include retail and entrepreneurial contexts, and eight additional lab sets have been built for non-retail learners.

Build on what exists

📊

A structured evidence base for iteration

Three cohorts produced 203 pieces of learner feedback, 57 baseline profiles, three cohort-level dashboards, and detailed instructor observations from every session. The patterns across those three datasets are now clearly documented. The program knows which modules consistently outperform, which delivery conditions produce the strongest engagement, which facilitation practices work best, and where the gaps are. Future cohorts benefit from that accumulated knowledge.

📄

Full intellectual property ownership

All curriculum, instructional materials, assessments, multimedia, and supporting documentation are owned outright by CILAR. The program can be updated, adapted, licensed, and scaled without restriction. Content can be repurposed for onboarding programs, HR training, professional development initiatives, or any other learning context where AI literacy is relevant.

🔧

Curriculum revisions and facilitation standards in progress

A structured round of curriculum revisions is underway across all eight modules, incorporating what the pilot cohorts taught. Career-similarity breakout grouping, which produced measurably stronger engagement in Cohort 2, is already codified as a standing facilitation principle. Module 7 is being updated to offer a structured individual-and-group blend. Module 8 content is under review. These changes go into all future cohort deliveries.

What the data shows

Framework-based learning transferred to real work. The AI Decision-Making Matrix, DECIDE Check, AI Responsibility Gate, and prompt engineering framework were each cited independently by learners across all three cohorts as standout takeaways. In every cohort, at least one learner reported applying a framework to a real situation at work before the program concluded. That is the clearest possible evidence of transfer.

The program produced genuine confidence growth. Across all three cohorts where pre- and post-program confidence data was collected, AI confidence grew from an average of 6.3 to 8.3 out of 10 (a combined average gain of approximately 2.0 points). Qualitative reflections across all three cohorts confirmed the same pattern: learners arrived with varying levels of familiarity and left with more structure, more intention, and more confidence in their judgment.

Cohort 2 demonstrated what the program looks like at its best. At 84% average attendance, 4.49/5 engagement, 89% usefulness ratings of 4 or 5, and an AI confidence gain of 2.65 points, Cohort 2 set a clear benchmark. The conditions behind that result — a cohort with shared professional context, deliberate career-similarity grouping in breakout rooms, and experienced facilitation — are now documented and replicable.

Knowledge retention stayed strong even when attendance and engagement varied. Cohort 3's knowledge check results (96% mid-program, 84% end-of-program) were the highest of the three cohorts, despite its lower attendance average. The content worked for the learners who showed up and engaged with it.

Module 6 was a consistent standout. Across all three cohorts, Module 6 (adapting AI skills across roles) was among the highest-rated sessions of the program. The format — a scenario activity with four structured questions that applied every prior framework — generated the strongest group discussions and the most consistent positive instructor observations of any module in the pilot.

Self-paced completion has a consistent late-program gap. The same declining pattern appeared in all three cohorts: strong completion in Modules 1 and 2, gradual decline through the middle, and low completion for Modules 7 and 8 where assessment scaffolding is most important. This is the pilot's most consistent structural finding and the one with the most direct bearing on microcredential readiness.

Final-week attendance declined in two of three cohorts. Weeks 7 and 8 carried the microcredential preparation lab and program wrap-up, yet posted the lowest attendance of the program in Cohorts 1 and 3. Communication strategy, clearer framing of the final sessions' stakes, and a more compelling close to Week 6 are already in planning for future cohorts.

The program is built for broader deployment. Three cohorts built something larger than three cohorts: a complete content library, a proven facilitation approach, a sustainability infrastructure, and an evidence base. The Train-the-Trainer series, Program-in-a-Box, and On Demand course mean the program can now reach learners at any scale, in any format, with CILAR running it independently. The pilot was the foundation. What comes next is the program at full reach.