Not another AI module
STEP comes after AI literacy: students must show judgment, not just that they know the tools.
Singapore context
After AI literacy, students need workplace judgment.
Singapore is moving AI competencies into higher education from 2027. STEP helps universities answer what comes after: can students use AI safely, check it, explain it, and deliver work a manager would trust?
For deans, faculty leads, and curriculum teams preparing students for trusted AI-assisted work.
Where STEP fits
AI literacy is the floor. Workplace trust is what comes next.
AI courses teach students about tools. STEP tests whether students can use those tools responsibly when the brief is unclear, the AI output is imperfect, and the stakes are real.
Trust before day one
Students practise the checks and explanations that help managers trust new graduates faster.
Evidence leadership can review
The cohort report shows where students earned trust and where judgment broke down, with enough detail for faculty, careers, and boards to act on.
AI learning model
STEP focuses on the two layers closest to workplace trust.
The Four Learns give universities a simple way to place STEP beside existing AI courses without turning it into another content class.
Layer
Typical university focus
STEP contribution
How STEP tests it
Learning About AI
What AI is, how it works, and its limits.
Assumed baseline before the capstone.
Baseline
Learning to Use AI
Prompting, tool use, and field-specific application.
Assumed baseline; students may use AI inside missions.
Baseline
Learning With AI
Using AI to deepen work, not avoid thinking.
Students must frame the task, inspect sources, and improve AI output.
STEP focus Source-first checks before AI help becomes useful.
Learning Beyond AI
Judgment, responsibility, safety, and human accountability.
Students explain what they trusted, changed, and kept under human control.
STEP focus Explanation before the student submits final work.
Simulation design
AI is available, but the student still has to think.
STEP protects the struggle of learning. Students cannot simply accept a polished AI answer. They must check the notes, find weak claims, protect privacy, and explain the judgment behind the final work.
In every mission, AI offers a shortcut that looks good but is not safe. Students must resist the shortcut and earn the right answer.
1
Read the work context
Understand the manager, the audience, the deadline, and what could go wrong.
2
Check the facts
Separate verified notes from AI claims, guesses, and missing information.
3
Set AI boundaries
Decide what AI may help with and what must stay under human judgment.
4
Deliver and explain
Send useful work and explain the choices clearly enough for a manager to trust.
University outputs
One cohort run creates evidence for several teams.
One cohort run produces records that career services, faculty, and leadership each use differently.
Career and transcript evidence
Readiness records show what the student did, which skills were tested, and where they are ready to speak with confidence.
Teaching signal
The cohort view shows where students overtrusted AI, missed important checks, or need more practice before work.
Board-ready evidence
The report gives a practical readiness story, with strengths, gaps, and next steps for faculty review or governance reporting.