Homework no longer measures learning
AI can solve many take-home tasks. AINP shifts the target from outputs to traceable reasoning and verification.
AINP — AI-Native Pedagogy
Designing university courses where AI is part of the architecture of learning.
A structured framework for interaction, assessment, and course design in the age of generative AI— focused on process visibility, verification, and longitudinal evidence.
AINP
AI-Native Pedagogy
AICA
AI-Native Course Architecture
L4
Portfolio
Longitudinal, auditable evidence of growth
L3
Assessment
Reasoning trajectories; verification over answers
L2
Studio
Guided exploration, discourse, and micro-challenges
L1
Interaction
Transparent AI use: logs, critique, validation
Studios
Portfolio
Assessment
Certification
Framework-first launch: legitimacy + structure before tools.
Higher education needs new architecture
Generative AI breaks the assumptions behind traditional assignments and assessment. AINP provides a defensible structure.
Assessment integrity is under pressure
Instead of bans or loopholes, AINP makes AI use explicit and auditable through required artifacts.
Faculty lack usable models
Most guidance is tool- or policy-centric. AINP is a course architecture methodology faculty can adopt and defend.
AINP framework
AI-native pedagogy treats AI as a first-class participant in learning. The framework is designed to be transparent, rigorous, and scalable.
Process visibility
Students document decision-making and iteration: prompts, revisions, rationale, and checks—so learning is observable.
Verification as a habit
AI outputs are hypotheses. Students must validate with tests, references, counterexamples, or alternative methods.
Judgment over answers
Assessment focuses on critique, tradeoffs, and decision quality—skills AI cannot outsource responsibly.
Longitudinal evidence
Portfolios capture growth across iterations, enabling authentic assessment aligned with real practice.
AINP principles
- AI is explicit, not hidden (disclosure as a learning behavior)
- Process is a first-class outcome (traceable reasoning)
- Verification is mandatory (AI outputs require checking)
- Assessment targets judgment (critique, validate, defend)
- Evidence is longitudinal (portfolio proves growth)
- Constraints are designed (clear modes: A0/A1/A2)
- Equity is engineered (scaffolds, access, norms)
AICA: AI-Native Course Architecture
AICA is the flagship architectural model inside AINP—designed for course redesign, defensible assessment, and repeatable adoption.
Layer model
L1 Interaction → logs, critique, validation. L2 Studio → guided inquiry. L3 Assessment → reasoning trajectories. L4 Portfolio → growth evidence.
L4Portfolio
L3Assessment
L2Studio
L1Interaction
AI use modes
AINP defines clear modes so expectations are explicit and defensible across assignments and assessments.
- A0 — AI-restricted (independent performance evidence)
- A1 — AI-assisted (transparent logs + verification)
- A2 — AI-collaborative (critique + alternatives required)
- A3 — AI-delegated (exploration only; not valid assessment evidence)
The AICA promise
Replace brittle “ban AI” policies with an architecture that makes learning observable, verifiable, and assessable.
Certification pathway
A scalable adoption model for faculty and departments. (You can launch this as soon as the templates + rubric are ready.)
1
AINP-Ready
Policy + artifacts + rubric defined.
2
AINP-Implemented
At least one full studio cycle + portfolio entry.
3
AINP-Verified
Evidence aligns outcomes, integrity, and assessment.
4
AINP-Exemplary
Reusable kit + exemplars + outcomes narrative.
Interested in piloting AINP?
Start with the canonical templates and a single studio cycle. You can expand to certification when ready.
Resources
Canonical artifacts to make AI use transparent and assessment defensible. (These are placeholders—link to PDFs or docs when you publish them.)
AIL Template
AI Interaction Log for transparent AI use (prompts, critique, verification).
Add link →Syllabus Policy
Defensible AI policy language aligned with AINP/AICA modes.
Add link →Portfolio Template
Evidence-first portfolio entry structure for authentic assessment.
Add link →Studio Cycle
Explore → Constrain → Validate → Reflect cycle with AI modes.
Add link →Whitepaper
AINP is designed to be publishable. Launch credibility comes from a clear problem statement, definitions, and canonical artifacts.
What the whitepaper should include
- Why legacy assessment breaks under generative AI
- Definitions: AINP, AICA, AI use modes
- Canonical artifacts (logs, rubrics, portfolio)
- Implementation patterns + case examples
- Equity and integrity considerations
- Evaluation plan and outcomes measures
Publishable framing
Position AINP as an architectural framework aligned with studio-based learning, metacognition, authentic assessment, and cognitive apprenticeship—extended for AI-mediated learning.
Replace this block with your actual PDF link when ready.
Add whitepaper linkAbout
AINP is built from classroom-tested implementations and research-informed course design.
Positioning
AINP is framework-first: it prioritizes legitimacy, defensibility, and adoption by faculty and institutions. Tools and platforms can be built later on top of validated methodology.
- Audience: STEM faculty, departments, faculty development programs
- Focus: assessment redesign, interaction structure, portfolio evidence
- Outcome: auditable learning in an AI-rich environment
Contact
Add your preferred contact method here (email form, address, or institutional page).