School-based Policy in Action: Operational Implementation Guide

Phase 1: Compliance, Privacy & Risk Governance

• Ban PII input: Prohibit staff and students from entering Personally Identifiable Information (PII)—including student names, assessment records, attendance logs or behavioural histories—into public AI platforms.
• Apply proportional risk zoning: Categorise software using structured risk models (such as the VINE 3-Risk-Zone framework): Zone 1 for personal teacher productivity, Zone 2 for vetted classroom tools and Zone 3 for non-identifying student data requiring formal Privacy Impact Assessments.
• Verify tool safety via ST4S: Check security ratings on Education Services Australia’s Safer Technologies 4 Schools (ST4S) portal before procuring EdTech to confirm student data is not harvested or repurposed for commercial model training.
• Preserve human accountability and consent: Secure explicit opt-in parental consent for platforms requiring student accounts; ensure educators retain sole authority over final grading, official reporting and parent communications.

Phase 2: Protecting Cognitive Development & Curriculum Integration

• Prevent cognitive offloading: Design units that preserve 'productive struggle' so students do not use AI summaries or generated text to bypass close reading, decoding and deep conceptual synthesis.
• Distinguish scaffolding from offloading: Use AI to build background knowledge sets or decodable fluency scaffolds without trapping learners in permanent low-level texts that limit exposure to grade-level syntax.
• Embed ACARA v9 AI literacy: Audit unit plans against ACARA’s Curriculum Connection: Artificial Intelligence to teach algorithms, machine learning and data analysis explicitly across Mathematics, Technologies and Science.
• Target cognitive load & ZPD: Deploy adaptive platforms to monitor learner demand, triggering targeted support only when academic struggle becomes counterproductive.

Phase 3: AI-Resilient Assessment & Academic Integrity

• Reject unreliable AI detectors: Cease reliance on automated AI detection tools due to high error rates and documented false-positive bias against EAL/D and LBOTE students.
• Embed process checkpoints: Anchor assessment in observable milestones, including sequential draft portfolios, literature mapping, source triage and face-to-face oral defences.
• Increase supervised assessment: Re-establish supervised, in-class checkpoints and handwritten components for high-stakes summative evaluations in line with state authority directives.
• Mandate student integrity training: Require secondary students to complete structured academic integrity modules (such as QCAA's mandatory certification) to embed attribution rules and ethical habits.

Phase 4: Classroom Pedagogy & Ethical Practice

• Cultivate critical evaluation: Train students to interrogate algorithmic outputs for bias, hallucinations and omitted viewpoints rather than treating machine text as neutral fact.
• Prioritise high-impact feedback: Delegate routine grammar and surface mechanics to automated tools so teacher time remains focused on argument depth, voice and conceptual nuance.
• Protect Indigenous data sovereignty & ICIP: Strictly uphold Indigenous Cultural and Intellectual Property (ICIP) protocols; ensure First Nations cultural material, oral histories and artwork are never fed into commercial AI datasets without explicit authority.

National Infrastructure Support Matrix