Most AI professional development for teachers is a one-off webinar that changes nothing. Here are five strategies for building AI-ready teachers, with practical steps school leaders can start this term.
It's August, every teacher in the school district sits through a ninety-minute webinar called something like "Introduction to AI in the Classroom." A consultant demos three tools and slides are shared. A few days later, a certificate of completion lands in everyone's inbox.
By October, while a handful of early adopters are using AI daily, most teachers have not opened the tools since, and the teacher who was anxious about AI in August is now anxious AND behind. The district, meanwhile, reports that 100 percent of staff have been trained.
This is the familiar story behind almost every district's standard AI PD rollout.
The problem is that, on paper, AI professional development for teachers is booming. In the EdWeek Research Center's fall 2025 survey, 50 percent of teachers said they had received at least one PD session on using AI in their work, up from 42 percent the year before and just 13 percent in 2023.
A nationally representative district survey from RAND found the same pattern, roughly half of the districts reported that they had provided Gen AI training to their teachers.
So, if the lack of training is not the issue, the question is no longer whether to offer AI training. A better question that superintendents and principals could ask themselves, is, why so much of the training offered so far has not changed what happens in classrooms.
Today's blog will discuss five strategies that address that gap directly.
Strategy 1: Put Teachers in the Student Seat First
In a world where AI is quite literally on our faces everywhere we go, an AI PD session which opens with a presentation about AI would hardly be useful to the teachers. Effective AI PD for teachers should open with teachers using AI on their own real work, before anyone lectures them about it.
The reason is simple. A teacher who has spent twenty minutes drafting a parent email with an AI assistant, arguing with it, correcting its tone, and rejecting half its suggestions has learned more about the technology's strengths and limits than any slide deck can teach.
She has also had the two experiences every teacher needs before setting AI expectations for students, differentiating between the moment the tool saved her real time, and the moment it confidently got something wrong.
So, in practice, the first session should be a working session in a safe, low-stakes environment where nothing a teacher tries can break anything or embarrass them.
The best thing to do here would be to give the teachers a task they already have to do this week, such as differentiating a lesson, drafting a rubric, or writing a newsletter. Let them do it with an AI tool, in pairs, with permission to be unimpressed. Close by comparing what the tool did well and where it was not very useful.
This is the same principle behind TomoClub's game-based approach with students: nobody learns a skill by watching someone describe it. Teachers build teacher AI literacy the way students build collaboration in a game, by playing, failing safely, and trying again.
Strategy 2: Trade the One-Off Webinar for a Cycle
The biggest flaw in current AI PD is its shape. One session, however good, cannot build a durable practice, and one session is what most teachers get.
A cycle looks different. An initial hands-on session is followed, two or three weeks later, by a structured conversation about what teachers actually tried in their classrooms. This helps diagnose what worked, and what flopped. Teachers then set the next thing they will test. A feedback loop is what will actually help integrate AI into the teacher's daily classroom activities.
Districts rarely need new infrastructure for this. Most already run professional learning communities, and AI fits naturally into that rhythm as a standing agenda item: one teacher demonstrates one use, colleagues critique it, everyone leaves with something to try.
Between sessions, teachers also need a safe space to ask questions when something breaks on a Tuesday afternoon. A shared channel where colleagues and facilitators answer in real time, a running library of subject-specific templates, and a short weekly digest of what actually changed in the AI tools they use are what turn a training event into a practice.
Every TomoClub PD engagement includes this layer for exactly that reason, because the one-and-done problem is not solved by a better single session.
Strategy 3: Teach Judgment Through the Tools, Not the Tools Alone
Any PD built purely around the features of one product has a shelf life of about one update. Interfaces change, tools get acquired, and carefully produced tutorial screenshots stop matching what teachers see on screen.
That is not an argument for keeping PD abstract. Teachers learn AI by working inside real tools, and a session that never really gives them practice with that won't be very useful.
The distinction lies in what the tool time is for.
When teachers use AI to pull answers out of a stack of curriculum documents, the real lesson is learning when to trust what it gives back and when to push on it. When they chain a few prompts together to automate a routine task, the lesson is deciding where a human still needs to check the work.
Giving your staff a shared framework for student use, such as the 0-to-3 AI usage scale from the North Carolina Department of Public Instruction, where every assignment carries an explicit expectation from AI-free to AI-empowered can also be a useful way to figure the teacher's judgement in making the calls about how AI is used in classrooms.
Strategy 4: Differentiate PD the Way You Would a Classroom
No teacher would run a single lesson plan for students at wildly different styles of learning and capabilities and call it done. But your average AI PD does this routinely.
Just like students, the range in any faculty is quite wide. In an EdWeek Research Center survey from early 2026, 13 percent of teachers said they were not at all eager to learn about integrating AI into their practice, and another 24 percent were only slightly eager.
Educators overall are close to evenly split on whether AI will be good or bad for teaching over the next five years. Sitting beside those colleagues, in the same room, is a teacher who has been building custom chatbots since 2023.
A differentiated AI PD programme would work because it takes both seriously. Start with a readiness survey so the content is built around where your staff actually are rather than a generic assumption about their preferences.
Early adopters can get more advanced work on research workflows, agents, and building simple classroom tools, plus a role piloting and teaching peers, whereas, their colleagues who are sceptical get something more valuable than persuasion, which is honesty. Their questions about data privacy, bias, and de-skilling should be a part of the agenda rather than being managed around.
RAND's researchers put this directly – teachers' fears about AI are common among adopters and non-adopters alike, and leaders should discuss them openly instead of minimising them.
Running distinct levels, with attendance by choice rather than assignment, keeps the beginner in low-stakes practice and the advanced user building something, feeling heard, engaged, and important.
Strategy 5: Start With The Most Time-Consuming Tasks
Teachers do not have a motivation problem with AI. They have a workload problem with everything, and PD that adds to the pile will be resented however good it is.
The fastest way to earn a faculty's attention is to aim the first months at the repetitive, time-consuming tasks that teachers most want off their plates – first drafts of lesson plans and rubrics, differentiation of existing materials, routine communication, and feedback on low-stakes work.
Teachers who use AI tools weekly report saving several hours a week, which compounds into weeks of reclaimed time across a school year. The district's most convincing promise, thus, would be time returned, not immediate adaptation and innovation.
Then measure the right thing. Completion certificates measure attendance. What leaders should look at is whether anything changed, so track what a cycle-based programme naturally produces – how many teachers tried something new since the last session, what they kept, what they abandoned, and whether teachers report spending less time on paperwork over a semester.
Across TomoClub's PD cohorts, 92 percent of participating educators report a stronger understanding of AI and 86 percent report being better able to identify which tasks are suited to it, which is the kind of baseline-to-outcome movement a district can put in front of a board.
When the first cohort can say the training gave them their Sunday evenings back, recruitment for the second cohort will take care of itself.
Where This Leaves School Leaders
The gap between districts is no longer a gap in access to AI tools. The tools are already in teachers' inboxes, embedded in the software they use daily. The gap is whether teachers have been given real practice, real repetition, and real honesty about the technology, or a webinar and a certificate.
The one principle underlying all the five strategies above is simple – teachers become AI-ready teachers the same way students become anything-ready, through guided practice over time rather than being told about a thing once.
Districts that build on that principle spend the next few years compounding small classroom wins. Districts that keep buying webinars keep reporting full training rates while their classrooms stay exactly the same.
TomoClub's AI Fundamentals PD is built as practice rather than presentation: hands-on, customised to your school's context after a readiness survey, and backed by ongoing coaching and community support once the session ends. Book one today.
Frequently Asked Questions
What is AI professional development for teachers?
AI professional development for teachers is structured training that helps educators understand what AI is, use it competently in their daily work, and guide student use responsibly. Strong programmes cover four areas: foundational awareness, practical efficiency with lesson planning and administrative tasks, ethics and academic integrity, and personalisation for diverse learners. The distinguishing feature of effective AI training for educators is that teachers spend most of the time working with the tools rather than watching a presentation about them.
How much AI training do teachers actually need?
There is no single answer, but the shape matters more than the hours. A half-day or full-day foundational session followed by ongoing coaching produces more classroom change than the same total hours delivered as scattered one-off webinars. Most districts see the strongest results when initial training is followed by structured check-ins every few weeks during the first term.
How do we handle teachers who are sceptical about AI?
Put their concerns on the agenda rather than around it. Research from RAND found that worries about privacy, bias, and the technology's role are common among AI adopters and non-adopters alike, and that leaders should discuss them openly. In practice, sceptical teachers often become the strongest voices for responsible use once their questions are treated as legitimate rather than as resistance to be overcome.
How do you measure whether AI professional development worked?
Attendance and completion rates measure neither learning nor change. Better measures include the proportion of teachers who tried a new AI use between sessions, self-reported confidence in prompting and tool selection before and after, time saved on planning and administrative work, and the number of classrooms with explicit AI expectations written into assignments.
Should teachers be trained before students start an AI literacy curriculum?
Yes. Teachers who have practised with the tools themselves can facilitate student sessions with confidence and answer the questions students actually ask. Running student AI integration in the classroom ahead of teacher training tends to produce inconsistent delivery and leaves teachers improvising in front of a class.