From Generative AI Experimentation to Better Lesson Design

Lesson planning requires teachers to balance curriculum expectations, learning progression, available time, classroom diversity and assessment demands. Generative AI can help produce initial drafts and alternatives, but effective teaching still depends on professional expertise.

This course focuses on using AI as a supervised planning assistant rather than an automated decision-maker. Participants explore workflows that begin with learning objectives and conclude with teacher-reviewed materials.

The training is suitable for primary, secondary, vocational and adult education professionals. Participants bring an example curriculum topic, unit or learning challenge from their own teaching context. Those who prefer not to share institutional materials can work with fictional or publicly available examples.

The opening workshop compares a generic AI-generated lesson plan with a carefully specified, curriculum-informed request. Teachers analyse differences in clarity, sequencing, inclusiveness and feasibility.

They then create their own prompt templates for future planning tasks.

Writing Prompts That Support Pedagogical Quality

Good AI-assisted planning begins with a clear instructional brief.

Participants learn a structured prompting method that includes the learners’ educational stage, prior knowledge, subject content, learning objectives, available resources, lesson duration and assessment evidence.

For example, a mathematics teacher may request a 45-minute lesson on proportional reasoning that includes an introductory misconception check, collaborative problem-solving, visual explanations and an exit ticket.

The generated plan is not accepted automatically. Participants identify unrealistic timing, unnecessary activities, unclear explanations and missing prerequisites.

Teachers then refine their prompts and compare outputs to understand how the quality of the instructional brief influences the usefulness of generated materials.

A reusable prompt library is developed for lesson sequences, explanations, learning tasks, examples and reflection activities.

Designing Coherent Lesson Sequences

Participants move beyond isolated activities to develop a connected teaching sequence.

They use backward design principles: identifying desired learning, deciding what evidence will demonstrate understanding and selecting activities that prepare learners to produce that evidence.

A science teacher might design a five-lesson sequence on ecosystems. AI may suggest an introductory investigation, vocabulary tasks, data interpretation, group discussion and a final explanation task.

The teacher checks scientific accuracy, curriculum alignment, progression and workload before revising the sequence.

Participants also explore how AI can propose alternative explanations, classroom examples and homework tasks while preserving the teacher’s original learning intentions.

The workshop concludes with a five-lesson plan containing objectives, activities, differentiation strategies and assessment checkpoints.

Differentiation Without Lowering Expectations

A central focus is adapting learning experiences for mixed-ability and multilingual classrooms.

Teachers examine how AI can help rewrite instructions in clearer language, create scaffolded tasks, generate worked examples and offer alternative ways to access complex content.

For example, in a history lesson about migration, participants might produce a simplified source-analysis guide, a vocabulary-supported version and an extension task requiring comparison of perspectives.

All versions retain the same central learning goal, while varying the support provided.

Participants check whether generated adaptations oversimplify concepts, introduce stereotypes or unintentionally reduce challenge for particular groups.

Accessibility is considered throughout. Teachers evaluate readability, document structure, alternative text, visual clarity and compatibility with assistive technologies.

Where AI tools are unavailable or unsuitable, the course explores equivalent teacher-created alternatives.

Designing Meaningful Formative Assessment

Generative AI can produce large quantities of questions, but quantity does not guarantee assessment quality.

Participants learn to create assessment tasks that reveal reasoning, misconceptions and progress towards specific objectives.

Workshops include developing diagnostic questions, exit tickets, short quizzes, discussion prompts and retrieval-practice activities.

Teachers examine distractors in multiple-choice questions, ambiguous wording and answers that may be technically correct but pedagogically misleading.

A language teacher, for example, might create a short writing task and a checklist for assessing coherence, vocabulary range and communicative purpose.

Participants revise generated assessment materials to ensure that they are fair, accessible and aligned with the learning being assessed.

The goal is to produce useful evidence of learning rather than simply automate question generation.

Rubrics, Feedback and Human Judgement

Teachers practise generating analytic rubrics with clear criteria and performance descriptors.

They compare vague descriptors such as “good understanding” with observable statements that specify what a learner can demonstrate.

Participants then test rubrics against fictional student work samples to identify inconsistencies and unclear expectations.

Feedback workshops explore how AI can help draft constructive comments that identify strengths, suggest next steps and encourage revision.

However, the course explicitly distinguishes feedback drafting from automated grading. Participants discuss the limitations of AI-generated evaluations and the importance of teacher responsibility, especially in consequential assessment decisions.

Teachers practise rewriting generic feedback into specific, respectful and actionable guidance.

Academic Integrity, Privacy and Copyright

AI-assisted assessment introduces questions about authorship, transparency and student independence.

Participants develop classroom guidance explaining when AI assistance is acceptable, when it must be acknowledged and when work should be completed independently.

They also examine alternatives to relying on AI-detection software, such as staged drafting, oral explanation, process portfolios and reflective learning logs.

Privacy workshops use fictional student profiles and synthetic work samples. Participants are instructed not to upload identifiable student records, grades or sensitive information to unapproved services.

Copyright discussions address source material, generated resources, licensing, attribution and the need to verify any reused content.

Building a Reusable Teaching Toolkit

During the final day, each teacher assembles a coherent instructional package.

The package includes a five-lesson sequence, differentiated resources, formative assessment activities, an analytic rubric, feedback templates and a prompt library.

Peer-review sessions use a quality assurance checklist covering accuracy, curriculum alignment, inclusion, feasibility, assessment validity and responsible AI use.

Participants revise their resources and create a short classroom implementation plan.

Classroom-ready takeaways: a reusable prompt bank, five-lesson teaching sequence, three differentiated activity versions, formative assessment set, analytic rubric, feedback templates and a responsible AI planning checklist.

The intended result is a more structured, critical and pedagogically sound approach to AI-assisted preparation, not dependence on AI-generated materials.