Skip to content

README.md

Latest commit

 

History

History
139 lines (88 loc) · 3.78 KB

README.md

File metadata and controls

139 lines (88 loc) · 3.78 KB
Raw

AI Integrity Design Tool

A local-first tool for designing clear, structured, and governance-ready approaches to acceptable AI use in assessment.

🌐 Live Hosted Version
http://cloudpedagogy-ai-integrity-design-tool.s3-website.eu-west-2.amazonaws.com/

🖼️ Screenshot
AI Integrity Design Tool Screenshot

🔗 Role in the CloudPedagogy Ecosystem

Phase: Phase 4 — Assessment & Integrity Layer

Role:
Supports the design of AI-aware academic integrity approaches by defining acceptable use, conditions, risks, and guidance for both students and staff.

Upstream Inputs:

  • Assessment structures from the Assessment Design Engine
  • Institutional policy context (optional)

Downstream Outputs:

  • Structured guidance for the Evidence & QA Pack Generator
  • Inputs for Curriculum Change Manager (tracking policy evolution)

Does NOT:

  • Detect or police misconduct
  • Replace institutional policy frameworks
  • Automatically generate compliance decisions

Additional Documentation

Overview

The AI Integrity Design Tool supports a shift from reactive enforcement to integrity-by-design.

It enables educators and programme teams to:

  • Define acceptable, unacceptable, and conditional AI use
  • Generate clear student-facing guidance
  • Document staff rationale and authenticity expectations
  • Identify risks and assumptions in AI-enabled assessment

This helps create assessment environments that are:

  • transparent
  • consistent
  • pedagogically aligned
  • governance-aware

Key Features

  • AI Usage Positioning
    Define whether AI use is allowed, restricted, or conditional

  • Structured Use Definitions
    Specify acceptable and unacceptable uses

  • Conditions & Constraints
    Capture boundaries for appropriate AI use

  • Guidance Generation
    Produce student-facing guidance and staff rationale

  • Risk & Authenticity Design
    Identify risks and authenticity considerations

Technical Overview

  • Built with TypeScript + Vite (React)
  • Fully local-first — runs entirely in the browser
  • Uses localStorage for persistence
  • Supports JSON import/export
  • No backend or external data storage

Run Locally

npm install
npm run dev

Build

npm run build

Design Principles

  • Local-first and inspectable
  • Governance-aware by design
  • Structured, not automated decision-making
  • Supports human judgement rather than replacing it

Disclaimer

This repository contains exploratory, framework-aligned tools developed for reflection, learning, and discussion.

These tools are provided as-is and are not production systems, audits, or compliance instruments. Outputs are indicative only and should be interpreted using professional judgement.

  • All applications run locally in the browser
  • No user data is collected, stored, or transmitted
  • All example data is synthetic and does not represent real institutions or programmes

Capability and Governance

This tool supports both AI capability development and lightweight governance.

  • Capability is developed through structured interaction with real workflows
  • Governance is supported through optional fields that make assumptions, risks, and decisions visible

All governance inputs are optional and designed to support — not constrain — professional judgement.

About CloudPedagogy

CloudPedagogy develops open, governance-credible tools for building confident, responsible AI capability across education, research, and public service.