The way you introduce AI to teachers will determine whether they use it or fear it

I have introduced AI tools to teachers in training sessions more times than I can count now. And I have learned something that the breathless AI-in-education conversation almost never acknowledges: the tool is almost never the problem. The introduction is.

When a teacher walks into a session and the first thing they see is a demonstration of what AI can do — generate a lesson plan in thirty seconds, write an assessment rubric, produce differentiated tasks for five ability levels — one of two things happens. Either they feel excited, or they feel afraid. And what determines which one? Not the tool. Not the teacher's age or experience or openness to change. It is almost entirely about whether the introduction starts with the teacher's world or with the technology's capabilities.

> Start with the technology and you are showing teachers what they might be replaced by. Start with the classroom and you are showing them what they might be freed from.

What I have seen in the room

I have sat with a room of teachers — experienced, credentialed, genuinely committed to their craft — and watched their faces shift from cautious curiosity to something close to defensiveness the moment AI started producing content that looked like their work. The implicit message landed before any word was spoken: this can do what you do, faster.

But I have also sat with the same profile of teachers and watched something completely different happen when the session started differently. When we began not with "look what AI can do" but with "what takes you the most time and gives you the least return" — the room changed. Teachers started talking about marking. About writing the same feedback comment for the fifteenth time. About producing differentiated materials for mixed-ability classes at ten o'clock on a Sunday night. And then, when AI was introduced as a possible answer to those specific frustrations — not as a capability showcase but as a targeted tool — the reaction was almost universally different.

This is not a subtle distinction. It is the entire difference between a teacher who leaves a session and experiments, and one who leaves and never opens the tool again.

The framework that actually works

After running these sessions across premier Cambridge-affiliated, ICSE, and CBSE schools in India, I have landed on a simple sequencing principle: problem first, tool second, judgment always.

Problem first. Name a real, specific, recurring friction point in the teacher's working week. Not a hypothetical. Not a generic challenge. Something that costs them time and energy regularly. Marking written work. Writing individualised reports. Preparing speaking practice tasks. The more specific, the better.

Tool second. Once the problem is named and the teacher owns it, introduce AI as one possible approach to that specific problem. Not a general demonstration. A targeted one. Show exactly how a tool handles that particular task. Let the teacher evaluate whether the output is any good.

Judgment always. This is the part most AI-in-education training skips. The teacher's professional judgment does not become less important when AI is in the room — it becomes more important. Someone has to decide whether the AI-generated lesson plan is actually appropriate for this class, on this day, with these learners. That someone is the teacher. Making this explicit changes how teachers relate to the tool entirely. They are not being replaced. They are being given a first draft that still needs their expertise to become anything worth using.

What this means for schools introducing AI

If your school is planning an AI introduction for teachers — whether that is a single workshop, a series of sessions, or an embedded programme — the single most important design decision is not which tools to use. It is where you start.

Starting with a demonstration of AI capabilities is tempting because it is impressive. But impressive is not the same as useful, and useful is not the same as adopted. The schools I have worked with that have seen genuine, sustained AI integration into teaching practice are the ones that started with teacher frustration, not technology fascination.

They also — and this is worth noting — gave teachers time to fail with the tools safely. To produce bad outputs and evaluate them. To realise that AI does not actually know their Year 8 class or their school's assessment rubric or the particular student who needs a slightly different approach. That realisation, experienced first-hand rather than told from a stage, is what builds the professional confidence to use AI with genuine judgment rather than either fearful avoidance or uncritical adoption.

> AI readiness for teachers is not a technology problem. It is a professional confidence problem that technology can either worsen or help — depending entirely on how it is introduced.

A note on what AI readiness actually means

I want to be specific about this because the term gets used loosely. AI readiness for a teacher is not knowing how to use ChatGPT or Gemini or any specific tool. Tools change. The underlying capability — the professional judgment to evaluate AI output critically, use it selectively, and take full responsibility for what goes into a classroom — that is what readiness actually means. And that capability is built through exactly the kind of contextual, problem-first introduction I have been describing. Not through a one-hour demonstration and a list of prompts to try.

This is what I work on with schools. Not AI adoption as an end in itself, but AI readiness as a dimension of professional development — one that prepares teachers not just for the tools that exist today but for the judgment they will need as those tools keep changing.

If you are a school leader thinking about how to introduce AI to your teaching staff, I am happy to have a conversation about what that looks like in practice. Reach me at denver@denverd.in or through the contact page at denverd.in.