Teachers already work 50-hour weeks. Here's how school leaders can build AI readiness that runs through existing systems instead of adding to the pile.
The new school session is round the corner, and the break brought a fresh wave of AI vendors and tools into your district. You want to stay ahead of it.
But the readiness plan you drafted over the summer meets a staff who have already started without it. Why the reluctance, you might wonder, and how do you bring them along?
According to the Walton Family Foundation and Gallup's Teaching for Tomorrow: Unlocking Six Weeks a Year With AI study, six in ten teachers used an AI tool for their work during the 2024-25 school year. Most of them got there on their own. Teachers were more likely to teach themselves how to use AI than to receive training from their school or district, and roughly two-thirds had no district-provided training at all.
Your teachers aren't the bottleneck. What's worth asking is what it costs them to figure this out alone, and whether your plan reduces that cost or adds to it.
RAND's State of the American Teacher survey puts the average teacher week at around 49 hours, with administrative work outside of teaching among the top reported sources of job stress. A mandate to use the newest tool lands on top of that.
Building AI readiness for teachers, then, is less about strategy and more about order: what comes first, and how it can fit routines that already exist.
1. Adoption Is Not Your Bottleneck
The self-teaching figure is a warning, not a win
Half of your teachers teaching themselves to use AI tools might look like initiative on paper. Fit it into a working day, though, and it means individuals evaluating things after school, double-checking unclear instructions, comparing tools, reading terms of service, and working out what a safe prompt looks like. Sounds messy? Imagine working through all of it. If forty teachers spend two hours each on these tasks, that is eighty hours of unpaid work with no consistency, no privacy review and no measurable outcome.
A district that stays out of this hasn't saved anything. It has moved the cost onto individuals, without assessing what is actually useful for supporting classroom work.
The Arithmetic Behind Teachers' Hesitation
A teacher who seems reluctant about the latest AI tool is hardly making a statement about technology. Their decision most probably rests on assessing the hours left in their term, then choosing instruction over experimentation, which, in their position, is the right call. Treating that as something to solve with a better demo doesn't change the arithmetic.
The Gallup data points the same way. Teachers who use AI weekly estimate saving 5.9 hours a week, roughly six weeks across a school year. Teachers who use it only monthly save half of that. These are teachers' own estimates rather than observed time studies, so take them with a pinch of salt. But the shape is clear enough: the benefit is real for people who get past the learning curve, and it scales with how much you use the tools. The curve itself is what takes a long time and it obviously won't fit in an overfull work week.
Readiness gets measured where the overwhelm isn't
You did take your precaution and prepared a readiness audit. It seems to cover it all – device and connectivity coverage, tool inventories, policy status, vendor compliance, and staff perception scores. All of this makes for a very good audit. However, one detail that it cannot tell you is if a teacher has hours to spare during the week, or will be overwhelmed trying to crunch it in.
Addressing this isn't impossible. You can add a measure which keeps track of the hours on one non-instructional task, from a sample of staff, before introducing a new tool or programme. Without it, a plan can score well on every indicator you have while the people expected to deliver it struggle to fit it in their schedule.
2. Find Out What's Already Full Before You Add Anything
Ask one question, in an already planned meeting
Before the tool, the committee or the policy, find out what is taking up your teachers' time. One question can do more work than you think, ask them – what takes up most of your time in a week that isn't direct instruction?
You don't need a separate instrument for this, or an AI-perception survey. Run it in five minutes at a staff meeting already on the calendar. What may come back will be a mix of grading, family communication, differentiation and documentation. That gives you a real pool of tasks contributing to workload, and tells you where the educators will adopt AI and most welcome the help.
Point the first tool where the returns are largest
Among teachers who already use AI in school tasks, the largest reported time savings come from making worksheets and assignments, building assessments, administrative work and preparing to teach. Whereas, analysing patterns in student data and one-on-one instruction were areas of low return.
What's worth noticing is the gap between where AI saves time and where teachers actually reach for it. Grading is one of the least common applications, with roughly one in six teachers using AI for it monthly. Yet among those who do, close to eight in ten say it saves them time. The tasks with the highest reported payoff are not the tasks with the highest uptake, which means some of your district's easiest wins are sitting unclaimed.
Put your own survey results next to that gap and start where the two overlap. Cases vary district to district, and mapping this against real teachers in your schools will always beat a national average.
Jeff Court, a superintendent in Saskatchewan we interviewed for the Education Hall, ran a version of this before anyone was calling it AI. Rather than accepting the administrative infrastructure he inherited, he brought technology in to automate repetitive administrative tasks: HR processes, leave tracking, seniority records, etc. The point wasn't just efficiency. Administrative staff who had spent hours on data entry got that time back for the conversations that actually mattered. He describes the priority as keeping human connection intact, which he achieved by first making sure the boring, repetitive tasks were the first to be automated. The same approach will prove helpful in building AI readiness for teachers.
Take something off before you put something on
If AI is a real priority this year, something else has to be given up. Cancel a monthly staff meeting and use the slot for AI training. Drop a weekly report the tool can now produce. Hold another initiative until spring.
Because if you announce a rollout while nothing else moves, teachers will stack it on top of everything they're already carrying, which will obviously affect adoption poorly.
There's a second job you can lift from them, and most districts miss it. The teacher who spends her evening comparing tools and working out safe prompts with nobody to check her conclusions is the one who will be completely burnt out in two weeks.
This burnout is highly avoidable, by running one small operation. You can ask your IT, curriculum and legal teams to review a handful of tools together against your student data privacy rules, and then publish a short approved list with a line on what each one is useful for. This will produce a credible document, prepared by those qualified to do so, that teachers can consult instead of spending their weekends guessing. It will also keep you in the loop about how student data in your school buildings gets used.
3. Sequence It So the First Win Pays for the Second
Breadth is what makes AI feel like an initiative. Depth in one place is what makes it feel like relief.
One department, one term, one task, one number
Run a pilot before you roll anything out across the district. Pick the task your survey flagged most often, choose one tool for it, one department to try it, and one term to run it for.
Ask the sample group to time themselves on the task for two weeks before they start, then again for two weeks once they've settled in. Those two numbers will be the most important things to come out of the pilot. Where a vendor's "reassuring" claim would fail, a maths teacher telling another maths teacher that writing assessments went from three hours to fifty minutes will definitely work. This way, instead of introducing multiple tools that overwhelm the teachers, you'd know what to keep and what to leave out.
Fold it into a routine that already runs
Rather than bringing in AI as a standalone initiative, try to fit it into the structure already in place, whether it is planning cycles, the professional learning calendar, or routine curriculum reviews. Oregon's Department of Education makes a version of this argument in its own guidance, recommending that districts align AI guidance to existing policy, mission and values rather than drafting something new from scratch.
Lauren Bolack, a principal we interviewed for the Education Hall, demonstrates this best. She introduced AI tools by building them into systems her teachers were already using rather than launching anything separate, and the results spoke for themselves.
Let the teachers who went first do the demonstrating
A colleague showing the workflow they genuinely use could be more persuasive than an administrator, and considerably more persuasive than a vendor. Find one teacher per department or grade band who has already been experimenting and using the AI tools well, and give them time and recognition rather than an instruction to "convert" those non-AI to pro-AI. Their interactions with colleagues sets a positive example, makes AI adoption feel less overwhelming and builds motivation.
4. Make Getting It Wrong Survivable
Put the permission in writing
A teacher with no slack will hardly try to experiment with the tool. If a mistake costs a re-taught lesson, an awkward parent email or a poor evaluation, declining to try would probably look like the most sensible option. Giving a written permission, stating plainly that using AI for planning, feedback and communication is allowed and encouraged, removes a barrier most leaders don't realise even exists.
There's evidence that formalising this shows up in hours rather than just in tone. Gallup found that teachers at schools with a clear AI policy were more likely to have used AI in the past year, and that those schools saw a dividend around 26% greater: roughly 2.3 hours saved per week per teacher, against 1.7 where no policy existed. A clearly laid out AI Policy for the school with room for mistakes and growth is what will encourage teachers to actually take up the tools.
Any AI Policy, however, is framed best in dialogue with the stakeholders. Brad Winterod, superintendent of Georgetown Exempted Village Schools in Ohio, took this seriously and convened a committee of teachers and staff from across the district and built an AI strategic plan with their inputs rather than producing one in vacuum and handing it down. When the policy feels like a collective conversation rather than orders one has to reluctantly follow, the buy-in would be much smoother for your educators.
Name the limits so the permitted space is real
As with permission, you should also be explicit about the limitations of AI – clearly lay down what AI should not do in your schools: final grade decisions without teacher review, communication about student mental health or discipline, basically anything that stands in for a judgement about an individual child. Holding clear limits is what earns you the place to advocate for AI everywhere else.
That distinction, between work that should be systematised and work that has to stay human, is the whole craft, and mastering it wins you half the game in your district's success in AI adoption. A simple rule of thumb to go by is that technology can support and enrich instruction, but it cannot replace effective teaching.
Where This Leaves You
AI-Readiness and a readiness plan are different things. A district can hold a vetted tool list, a written policy and a trained committee and still have teachers who have never once used the AI tools in an actual lesson.
AI can do much to relieve the constant workload that remains one of the biggest causes of teacher burnout, but not when it is imposed on top of an already packed schedule. Treating the teachers like partners whose demands, needs and time are respected, rather than subordinates who should follow instructions, is what makes way for a successful, and sustainable AI-rollout that lasts.
An AI Policy where tools and programs exist to supplement, rather than add onto or try to replace the teacher's work, is what you should aim for. The hours you can give back to the teachers this way creates space for the more human interventions that shape the atmosphere of a classroom. Additionally, encouraging the teachers to trust their judgements regarding AI usage will also develop this habit in the students – the ability to judge an AI output, catch it when it's confidently wrong, and know when not to reach for it at all, skills that will make them ready to take on the future independently.
If these tips seem helpful for you and you cannot wait to test them out in your district, our AI Fundamentals PD is designed to fit your school and fits right into the teachers' calendar that's already full instead of competing with it.
Bring TomoClub to your school. Book a call →
Frequently Asked Questions
What does AI readiness for teachers mean?
It is not restricted to training hours or tool access, but also considers whether a teacher has the time, the permission and the support to use AI alongside the work they're already doing.
How do we introduce AI without adding to teacher workload?
To begin with, find the most time-taking, non-instructional task in the teacher's work week, and introduce one tool to increase efficiency, keeping in mind their already existing routine. Additionally, leave them room to make mistakes and lay out the limits of AI use clearly in your school's AI Policy.
Which AI tools help with repetitive administrative tasks?
For worksheets, assessments, admin and lesson prep, a general-purpose assistant covers most of it, and much of that already sits inside the Google Workspace or Microsoft 365 licence your district might hold. Purpose-built education tools are worth it for narrower jobs, like differentiating a text across reading levels. Vet anything centrally for student data privacy before it reaches a classroom.
How long should an AI pilot run before we scale it?
Long enough to produce evidence an educator would believe. One department, one term, one tracked task, with a mid-point check so you can adjust. Scale on what the pilot findings point to.