A teacher can watch an AI demonstration, understand every step, and still have a lesson to finish before tomorrow. The unfinished piece may be small: a discussion question that gets one-word answers, directions students keep asking about, or an exit ticket that reveals very little. Bringing it into the training gives the teacher something useful to improve before returning to class.
Short online sessions can introduce a tool. Google describes its AI Educator Series as free, on-demand training, with badges teachers can earn in under 15 minutes. Schools can build on that introduction by giving teachers time to try an idea on their own materials and discuss what happened.
Organize the next session around a lesson teachers already plan to teach. Give them time to revise a small part of it, check the result with a colleague, and choose when to talk about how it worked in class.
Bring a lesson that needs attention
An invitation to “try AI for lesson planning” leaves a teacher with a lot of decisions. Start with a reading, assignment, or lesson already in place. What would make it more useful for students?
For one teacher, that might mean clearer directions. An AI tool could suggest a shorter version, but the teacher needs to check that it keeps every step students need. A colleague can read the revision as a student would: is it clear where to begin, which materials to use, and what to submit? There is no need to build a new unit when an assignment for Thursday already needs attention.
Check the suggestions together
Consider a fictional seventh-grade social studies lesson about a town choosing where to build a bridge. In the teacher-created reading, farmers ask for a shorter route to market and council members discuss flood risk. The teacher asks an AI tool for discussion questions that ask students to explain the decision using details from the reading.
Suppose one suggestion asks, “Why did the council ignore the farmers’ concerns?” It sounds reasonable, but the reading says the council considered those concerns. The teacher and a colleague revise it: “How did the farmers’ concerns and the flood risk affect the council’s decision? Use details from the reading to explain.”
Then they try answering it themselves. Can they find the information students will need? Does the question call for an explanation, or can students finish by copying a sentence? They also check that students can read and answer it with the support they normally use.
A facilitator can show this kind of checking before teachers try their own materials. An imperfect suggestion gives the group a chance to discuss why a question needs changing or should be dropped. Teachers can then try it with a colleague or coach there to help.
Use public or teacher-created materials that are safe to share with a school-approved tool. Keep student names, work, and confidential records out of the exercise, and save any later classroom notes in the school’s approved systems. If teachers do not yet have tool access, they can check a prepared example together without spending the session creating accounts.
Plan when to try it and talk it through
Before leaving, the teacher chooses when to use the revision and when to discuss it afterward. Use an existing planning period or department meeting so there is no extra meeting to arrange.
The coach’s help can be specific. A teacher trying clearer directions might ask the coach to notice which questions students still ask before starting. If a classroom visit is not possible, the teacher can jot down where students needed help. Those notes give them something to discuss beyond whether the teacher liked the tool.
The Education Endowment Foundation’s professional development guidance recommends making a plan, getting feedback, and trying a technique again. UNESCO’s AI competency framework for teachers similarly recommends planning a lesson, teaching it, reflecting on what happened, and revising it, with help from coaches and colleagues. Trying the change in class gives teachers something to discuss and improve together.
Let the next conversation change the plan
What happens in class may reveal an unexpected difficulty. In the bridge example, students might identify both concerns but struggle to explain how they affected the decision. Showing students how to work through that reasoning could help more than another set of questions.
A teacher revising an exit ticket may find something similar. Students might correctly define a term without showing whether they can apply it. The next version could ask them to choose between two examples and explain their choice. Looking at the responses together helps the teacher and coach decide what to teach next.
Preparation time matters too. Count the time spent getting suggestions, checking them, making corrections, and adding the final material to the lesson. A quick AI response may take longer to fix than writing the question yourself. A suggestion may still be worth keeping if it improves the lesson without saving much time. Both the result and the time it takes are worth discussing.
Keep the original material beside the revision, with a short note about what happened in class and what to change next. That record will help when teaching the lesson again. It can also help a facilitator choose the next training topic from difficulties teachers actually faced.
For a team still deciding what to try, the Five-Minute EdTech Friction Audit can help identify a task and consider what a new tool would replace. Once the task is chosen, set aside enough meeting time to try it and discuss the result.
The next AI session can begin with a question teachers brought back from class. They have a lesson to revisit and someone to help them decide what would make it better.
