1. Use AI to learn how much of your LAP and tasks can be completed or imitated by AI

• Task Preparation: Select a representative sample of your assessment task sheets (the student handouts). Convert these into a suitable format (e.g., PDF, editable text). 
• Platform Selection: Use a secure, privacy-compliant AI platform (e.g., your school’s approved tool or a private, non-training instance of a public platform). 
• Safeguarding Intellectual Property (IP): Avoid public, data-mining platforms. Ensure your data will not be used for future AI training or shared externally.  
• Submitting Tasks: Input your assessment task into the AI/automation tool. Use a prompt such as: “Complete this SACE [insert stage, insert subject] assessment task based on the provided requirements.” 
• Review AI Output: Mark the AI-generated responses using your subject’s performance standards. Identify which criteria or task components are easily replicated and which resist automation. 
• Document Findings: Note any aspects of your assessment that may need redesign to improve authenticity or verification. 
• Prompt Templates and Privacy Considerations: 
• Provide clear guidance to staff on selecting prompts that simulate student responses authentically.  
• Ensure prompts and data are handled according to school privacy policies. Avoid uploading identifiable school or student information. 

Back to top

2. Update

• After conducting an automation check, you will have valuable insights into how easily your assessment tasks (or parts of them) can be completed or imitated by AI tools. You may also discover that some performance standards—such as factual recall or formulaic writing—are more susceptible to automation, while others—requiring personal reflection, unique problem-solving, or process evidence—are less so. 
• This information is powerful: it enables you to meaningfully adapt your Learning and Assessment Plan (LAP) and assessment tasks to better verify student learning, maintain academic integrity, and promote authentic achievement. Here are practical strategies, building on earlier sections of this toolkit, for how to respond: 

Identify Automatable Task Elements 

• List the Automatable Components: From your automation check, make a list of questions, prompts, or sections of tasks that the AI completed easily or at a high level. Note if certain performance standards are consistently “checked off” by generic, AI-generated responses. 
• Pinpoint Performance Standard Weak Spots: Identify which criteria are at the highest risk (e.g., ‘Factual Knowledge’) and which demonstrated less vulnerability (e.g., ‘Evaluation’, ‘Creative Application’, or ‘Personal Reflection’). 

Revise Tasks To Emphasize Authentic and Personalised Evidence 

• Integrate Process Evidence: Increase the value of drafts, reflections, annotated portfolios, or artefacts created during class or in supervised contexts.
• Embed Real-Time Interaction: Require elements such as live presentations, interviews, or teacher-led Q&A where students must explain and substantiate their thinking or process 
• Focus on Unique or Local Contexts: Reference class discussions, case studies, school projects, or recent learning experiences that can’t be anticipated by AI. 
• Negotiate Personalisation: Where possible, allow students to select topics or tasks relevant to their lives, interests, or local context—further increasing authenticity and uniqueness. 

Strengthen Assessment of Higher-Order Standards 

• Design for Analysis, Evaluation, and Creation: Prioritise tasks that require students to analyse scenarios, justify choices, adapt processes, reflect on their approach, or create original responses 
• Encourage Metacognition: Require written or oral reflection on how they learned or solved a problem, including challenges faced and strategies used. 
• Require Multi-Modal Evidence: Ask for combinations of visual, oral, and written evidence to demonstrate understanding beyond what an AI can easily generate. 

Apply Layered and Supervised Assessment Strategies 

• Break Tasks into Layers: Implement staggered checkpoints such as proposal, draft, peer review, and final product, each verified or observed by the teacher. 
• Supervised Milestones: Conduct parts of the assessment in class or in real time, such as planning, data gathering, or evaluation phases.

Back to top

3. Document and Communicate Adjustments 

• Record Revisions in the LAP: Clearly outline the changes made in response to the automation check, and update your LAP documentation accordingly. 
• Communicate Expectations to Students: Explicitly clarify what constitutes authentic evidence and the rationale for new assessment conditions or requirements. 

Back to top