Do it, review it, improve
Students act, see the record, and improve on a fresh task rather than memorising one answer.
After AI literacy: workplace judgment
Help graduates earn workplace trust before day one.
Graduates can move fast with AI. The harder question is: can a manager trust the work? STEP puts students in workplace situations where they must check AI output, explain decisions, and send work a manager would sign off on. Universities see who is ready, who needs help, and what to fix before graduates start work. AI modules teach students how to use the tools. STEP tests whether they can still think clearly when AI makes it easy not to.
- Evidence layer after AI literacy — tests judgment, not just tool use
- Aligned with Singapore's 2027 higher-education AI competency direction
- Student records for coaching and interviews · cohort reports for leadership
- Designed by AIR APAC. Developed by Exeter Labs.
Trust acceleration
Show your board what graduates can actually do with AI.
Singapore is moving AI competencies into higher education from 2027. STEP goes further: it shows whether students can use AI responsibly when real work is on the line.
Students produce real decisions, reviews, and recommendations, scored on the judgment skills managers look for but universities rarely test.
AI is available in every mission. But students must frame the problem, inspect the sources, and make a judgment call before AI helps. The work shows thinking, not just output.
STEP Campus runs as a 6-week program: test where students are now, work on weak areas, certify, and report results back to your team.
Workplace readiness task
The Trust Check
Open a simulated workplace, review an AI draft that looks right but is not, find the problem, and write a response your manager can trust.
Realistic pressure
6–8 min
Sample mission
The trust model
Trust is earned when the work removes doubt.
The test is not whether a student can use AI. It is whether they can protect the team from hidden risk while still getting the job done.
ContextSet up the work right
+
CheckTest what AI gave you
+
ExplainSay what you did
+
DeliverSend usable work
= Workplace trust
Defensible method
A serious method, shaped by real workplace pressure.
The design is based on proven assessment methods, but every scenario comes from real situations where new employees either earn trust or cause problems. Faculty can inspect the scoring guide, evidence records, and what the credential does and does not claim.
Work products, not recall
Scores come from real work products: reviews, recommendations, and decisions, not memorised answers.
Judgment with unclear information
Students make decisions with unclear instructions, realistic AI mistakes, and no perfect answer.
Multiple evidence points
The credential rests on several observable actions, not a self-report or quiz.
Claims tied to proof
Each task is built from the claim, the proof required, and the scenario moment that can reveal it.
Evidence before bigger claims
STEP claims only what it can show: how a student performed in the simulation, scored using a standard you can inspect.
Humans at the helm
Students must show what they trusted, checked, changed, explained, and kept under human control.
The next step after AI modules
AI courses teach how tools work. STEP tests whether students can use those tools safely when workplace trust is at stake.
AI available, not automatic
Students must frame the task, inspect sources, and make a judgment call before AI help opens. The work shows thinking, not just output.
Issued by AIR APAC
The credential is issued by AIR APAC using the STEP standard; Exeter Labs develops the platform.
Cohort run
From one cohort to decisions you can act on.
Start with one graduating cohort and one practical question: where are our students ready for AI-assisted work, and where will they need help before a manager trusts them?
-
01
Define what you want to learn
Pick the student group, the question to answer, and who needs the results.
-
02
Run the capstone
Students complete AI workplace missions under realistic time pressure and with incomplete information.
-
03
Score with a clear standard
Student work is scored using the STEP standard. Scoring guides and example answers help keep results fair.
-
04
Share the findings
Faculty, careers, and leadership see strengths, trust risks, and next moves.
-
05
Act and reassess
Students who need more practice work on weak areas, then try a new scenario.
The standard
Four behaviours managers notice fast.
Boards do not need another slide saying AI has been added to the curriculum. They need proof that graduates can think, check, explain, and deliver under pressure.
S
Structure
Turns an unclear request into a sensible plan before rushing into AI output.
T
Test
Catches AI mistakes that look correct before they become a manager's problem.
E
Engage
Explains their thinking, is honest about AI use, and handles tough questions without hiding behind the tool.
P
Perform
Delivers usable work when time is short, information is incomplete, and hard choices are real.
Structure, Test, Engage, and Perform give faculty, careers, and leadership one language for AI-era workplace judgment.
Quality standard
Rigor without a black box.
Universities can review the rubric, scoring guidance, results review process, and claim limits before a cohort launches.
Versioned standard
The standard names the skills, performance levels, and pass requirements behind the credential.
Consistent scoring
Scoring guides and example answers help keep marking consistent across cohorts.
Faculty results review
Results are reviewed with academic and careers teams, not dropped as a black-box score.
Claims stay specific
STEP shows how students performed in the simulation. It does not claim to predict job success or meet accreditation rules.
Academic reviewable
The rubric, scoring records, and group results can be reviewed by faculty or university panels.
AIR APAC credentialing
AIR APAC issues STEP certification; Exeter Labs provides the simulation and reporting platform.
Student outcome
A readiness record students can use.
When a student meets the standard, the record explains what they did, in which scenario, and which standard was used. It gives career teams better coaching language and students a stronger interview story.
STEP Certified: AI-Ready Graduate
Gives students a clear way to describe AI workplace skills beyond a course grade.
A record of demonstrated work
Shows the outcome, scenario, assessment date, standard version, and how the student performed across each skill.
Evidence a manager can understand
Explains the scenario, rubric, decisions, and work samples behind the claim.
Useful in the next conversation
Built for CVs, LinkedIn, interview prep, and career-services coaching.
Evidence for skills transcripts
The readiness record shows what the student did, how they scored, and which workplace skills were tested, ready for career portfolios and student records.
Growth on a fresh scenario
A fresh scenario shows improvement, not memorisation.
See the full record
Start with the sample mission, then inspect the student record and cohort flow.
Why STEP
Generic AI training stops before the hard part.
Prompting is easy to teach. The harder question is whether a graduate can check AI work, protect privacy, handle pressure, and explain decisions. Those are the things that earn trust in real workplaces.
Shows tools, workflows, and prompts.
Tests AI use inside difficult workplace situations.
Confirms completion or attendance.
Requires scored work samples and a clear pass.
Counts participation.
Shows where trust breaks down and what to improve next.
Student gets AI output and submits AI output.
Student must think first, then decide what AI should and should not do.
Institutional output
What a university receives.
One run creates the materials different teams need: a board brief, student records, a capability breakdown, and clear priorities for coaching or curriculum change.
Summary report for leadership
Cohort strengths, risk patterns, and practical next steps for graduate outcomes.
View sample reportIndividual student records
A per-student record for coaching, interviews, and employability portfolios.
Capability breakdown
Performance broken down into Structure, Test, Engage, and Perform.
Program quality evidence
Rubric details, group score breakdowns, and work summaries for quality reviews.
See evidence formatResults review with faculty and careers teams
A working session on what the cohort showed and what to prioritise next term.
Privacy by design
The public credential verifies the outcome. Student work stays governed by the institution and learner consent. See Privacy & data.
University teams
Careers, faculty, leadership. One shared picture.
Each team sees the same student work through the lens of the decision it has to make.
Student career evidence
Help students talk about their AI skills. Help coaches prepare them for interviews.
View sample cohort reportAn applied capstone with usable evidence
Run a workplace-focused final project without building a new assessment from scratch.
See how the capstone worksA cohort report for governance
See where students are ready, where judgment breaks down, and what to improve next.
View sample cohort reportA first run can start with one final-year or fresh-graduate group and end with student records, a group report, and a leadership review.