Prepare a training session with Claude: objectives, activities and assessment
By Dorian Laurenceau
Claude can help prepare a training session by organizing objectives, activities and materials. The trainer still chooses what learners should achieve and verifies the content. A polished lesson plan is not evidence that the session will work with your audience.
Our fictional example is a 90-minute adult-learning session on checking AI-generated answers. It requires no personally filmed videos: documents, discussion, diagrams and quizzes provide the learning activities. References were checked on October 14, 2026.
Define an observable learning objective
Replace “understand AI” with an action you can observe: “identify unsupported claims in a short response and justify two corrections using supplied sources.” State the audience, prior knowledge, accessibility needs and available equipment.
Tell Claude about the constraints. A session with shared computers, unreliable internet and mixed reading levels requires a different plan from an individual online course. Ask it to list assumptions instead of quietly assuming that every learner has an account and a laptop.
Design the assessment before the activities
Create a short fictional answer containing one correct claim, one unsupported claim and one misleading calculation. Label it explicitly as a teaching example rather than a real study or client report. Supply the source documents and an answer key.
The final task asks learners to annotate the answer, locate supporting passages and correct the calculation. Decide what counts as success before writing slides. This helps distinguish demonstrated reasoning from simply repeating the trainer’s explanation.
Build a realistic 90-minute sequence
| Stage | Time | Learner activity | Evidence of learning |
|---|---|---|---|
| Starting point | 10 minutes | Identify a claim that needs checking | Initial explanation |
| Worked example | 15 minutes | Compare an answer with its source | Located evidence |
| Pair exercise | 25 minutes | Annotate the fictional answer | Shared corrections |
| Group review | 20 minutes | Explain disagreements | Reasoned discussion |
| Individual task | 20 minutes | Apply the method to a new example | Annotated response |
These durations sum to 90 minutes. They are planning assumptions, not measured performance. Include instructions, transitions and the expected learner output in each stage. If the group needs more support, reduce the amount of material rather than compressing every activity unrealistically.
Produce materials and a correction guide
Ask Claude for separate learner and trainer documents. The learner version contains the task and sources; the trainer version adds expected observations, common mistakes and possible follow-up questions. Check that answers are not accidentally included in the learner handout.
Use concise diagrams to show the verification process and short quizzes to reveal misunderstandings. A quiz answer should explain why it is correct. It should not reward learners merely for memorizing a product name or repeating an absolute rule that has exceptions.
Review accuracy and accessibility
Open the referenced sources and check every important claim. Adjust vocabulary and document layout for the audience. Prepare an offline version of the materials so the session remains possible if an assistant or network is unavailable.
UNESCO’s guidance on generative AI in education provides a broader framework around human agency and appropriate use. It does not certify this lesson plan. Claude Academy is a source of product learning resources, not a substitute for your instructional judgment.
Draft your training sequence
Prepare your instruction here. This module sends no data to a model: copy your brief into your chosen tool.
View the starter instruction
Prepare a 90-minute session from the supplied audience, objective and constraints. Design the assessment first, then activities whose durations sum to 90 minutes. Specify learner outputs, instructions, materials, correction guidance and an offline alternative. Mark assumptions and do not invent research findings or student outcomes.
Improve after teaching
Collect the questions learners actually asked, the mistakes they made and the stages that took longer than expected. Compare these observations with the plan. Revise one part at a time and preserve the earlier version so the reason for each change remains visible.
Download the training preparation kit. For feedback on submitted work, use the human-reviewed rubric workflow.
Test Your Knowledge
0 / 3 answered
Q1.Which objective is observable?
3 questions remaining
Sources
Module 0 — Prompting Fundamentals
Build your first effective prompts from scratch with hands-on exercises.
Dorian Laurenceau
Full-Stack Developer & Learning DesignerFull-stack web developer and learning designer. I spent 4 years as a freelance full-stack developer and 4 years teaching React, JavaScript, HTML/CSS and WordPress to adult learners. Today I design learning paths in web development and AI, grounded in learning science. I founded learn-prompting.fr to make AI practical and accessible, and built the Bluff app to gamify political transparency.
One practical AI workflow in your inbox
Methods, resources and new exercises. Free, unsubscribe at any time.