Skip to content

AI professional development for teachers: a 90-day plan

AI professional development for teachers sticks when it lasts: 90 days of practice, coaching and evidence, not one training day. Here is the plan.

By Ernie Delgado6 min read
Illustration for: AI professional development for teachers: a 90-day plan

AI professional development for teachers changes classrooms when it lasts. Teachers need time to practice, get coaching, and show what changed, and one training day rarely allows that. Federal law agrees: its definition of professional development excludes stand-alone, one-day workshops.

Ongoing training is still rare. In the EdWeek Research Center's winter 2026 survey, only 9% of teachers called their generative AI training ongoing. This post lays out a 90-day plan to close that gap.

Why one-day AI PD doesn't change classrooms

Training is spreading fast. In 2023, just 13% of teachers told the EdWeek Research Center they had received even a little AI training. By fall 2025, half had attended at least one session.

Districts moved too. In RAND surveys, 48% of districts reported training teachers on AI by fall 2024. A year earlier, it was 23%.

But the depth has not kept up. Ongoing training is the exception, not the rule.

Teachers' AI training, two EdWeek Research Center surveys

No training, fall 2025 50% No training, winter 2026 42% Multiple sessions, fall 2025 14% Multiple sessions, winter 2026 22% Ongoing training, fall 2025 5% Ongoing training, winter 2026 9%
Source: EdWeek Research Center surveys of K-12 teachers: fall 2025 (training on AI) and winter 2026 (training on using generative AI in the classroom). Question wording differs between surveys, so compare with caution..

Gallup and the Walton Family Foundation surveyed public school teachers in early 2026. Only 18% had received any formal guidance on AI from their administrators. One workshop and no written rules leave a teacher with a tool and a guess.

The law already says this is not enough. The Elementary and Secondary Education Act defines professional development as "sustained (not stand-alone, 1-day, or short term workshops)." It also calls for PD that is intensive, collaborative, job-embedded, data-driven, and classroom-focused.

Effective PD is sustained, active, and coached

The research on teacher learning is older than ChatGPT, and it still holds. The Learning Policy Institute's 2017 review of 35 studies named seven features of effective professional development. It is content focused, uses active learning, and supports collaboration.

It also uses models of good practice, offers coaching and expert support, and builds in feedback and reflection. Finally, it is of sustained duration, so teachers have time to learn, practice, and refine.

Coaching has been studied closely. A 2018 meta-analysis by Matthew Kraft, David Blazar, and Dylan Hogan combined 60 studies with causal designs. The pooled effect of coaching was 0.49 standard deviations on instruction and 0.18 on student achievement.

The same study carries a warning. Effects in large programs were only a fraction of those in small ones. Scale is hard, so start with a plan your coaches can actually run.

A 90-day AI professional development plan for teachers

The plan below runs one school term. Each teacher picks one real, recurring task, such as a weekly quiz, a parent update, or leveled reading passages.

AI drafts. The teacher reviews, decides, and owns the result.

WeeksWhat teachers doWhat leaders check
1 to 2: Set the baselinePick one recurring task. Note how long it takes now. Learn the school's AI rules: approved tools, and no student data in prompts.Every teacher has named a task and a baseline. The approved tool list and rules are in writing.
3 to 4: First draftsDraft the task with an approved tool. Review and fix every draft before it reaches students. Log what you changed.Each teacher has a short AI log. Nothing went to students without review.
5 to 8: Coached practiceMeet a coach or peer every two weeks. Bring one AI-assisted draft and your edits to a team meeting.Coaching visits happened. Drafts show the teacher's edits, not raw AI output.
9 to 12: Classroom useUse the improved routine with students. Collect student work and quick feedback. Adjust.Student work samples are on file. No AI grades a student on its own.
13: Explain your workPresent the task: what AI did, what you changed, and whether to keep, fix, or stop it.Decide what spreads to other teams and what stops. Record the evidence.

Notice what teachers are doing in every row. They are building something they will use on Monday, not watching a demo. That is the same rule we hold for students: technology is a tool for creating, not consuming.

If a teacher can't explain what the AI did and what they changed, the training isn't finished.Ernie Delgado, BeyondK12

The weekly routine that makes it stick

The plan works only if the routine fits inside a normal week. Keep it small enough that a tired teacher can do it on a Tuesday.

  1. Pick the task. Choose the one task you are working on this term. Do not add a new tool every week.
  2. Draft with an approved tool. Use only tools on the school's list. Leave out student names, grades, and any other student data.
  3. Review and decide. Check facts, reading level, and fit for your students. Fix what is wrong, then make the call yourself.
  4. Log it. Write one line: the tool, the task, the time it took, and what you changed. This is your own AI disclosure record.
  5. Share one example. At the weekly team meeting, one teacher shows a draft and their edits. Colleagues ask questions.

This is the habit at the heart of our Explain your work series. Students in our TechPlus AI curriculum finish every project by explaining what they made, why, and which tools helped. Teachers should practice the same thing in PD, out loud, with colleagues.

How will you know in March whether October's PD worked?

Do not count attendance. A sign-in sheet tells you who was in the room, not what changed in the classroom. Count evidence instead.

If the 90 days start in October, week 13 lands in January. March is your follow-up check.

Are teachers still using the routine without being asked? Do their logs still show real edits? Has the student work improved?

Time matters too. Gallup found in 2025 that teachers who use AI at least weekly estimate they save 5.9 hours a week. Your own October baseline will tell you whether your teachers are anywhere near that, and on which tasks.

Coaching costs money. The U.S. Department of Education issued guidance in July 2025 on AI and federal grants. It says those funds can be used to train educators to use AI tools effectively and responsibly.

Title II, Part A funds professional development for teachers and principals, so it is the first place to look. Check with your program office before you budget.

Write the rules down before the first session, too. Teachers can't follow guidance that does not exist. Our guide to writing a school AI policy teachers will actually use covers what goes on that page.

One thing to do this week

Ask every teacher to name one task they want AI help with. Have them write down how long it takes today. That list is your baseline, and in March it shows whether October's training worked.

AI PD is a People question. To see where your staff stands, take the free AI Readiness Assessment. It scores your school across Policy, People, Curriculum, Guardrails, and Funding, so start with your People score.

If you want outside help with the coaching side, see our teacher professional development.

Related reading: Professional development for educators in the era of human and AI collaboration and Why school leaders must consider human in the loop AI.

Questions school leaders ask

How many teachers have had AI professional development?

About half had at least one session by fall 2025, according to the EdWeek Research Center. In its winter 2026 survey, 42% of teachers still reported no training on generative AI, 22% had multiple sessions, and 9% had ongoing training.

Why doesn't a one-day AI workshop work?

A single day can introduce a tool, but teachers need time to practice it on real work and get feedback. The Learning Policy Institute's 2017 review of 35 studies found effective PD is sustained, active, collaborative, and coached. Federal law's definition of professional development also excludes stand-alone, one-day workshops.

What should teachers do during AI professional development?

Pick one real task, draft it with an approved AI tool, review and fix the draft, and log what they changed. Then they bring the work to a team meeting and explain how AI helped. No student names or data go into any prompt.

Can federal funds pay for AI training for teachers?

They can, where the program allows. The U.S. Department of Education's July 2025 guidance says federal grant funds can be used to train educators to use AI tools effectively and responsibly. Title II, Part A, the federal program for teacher and principal professional development, is the usual place to start.

Is AI PD worth the time for busy teachers?

Gallup found in 2025 that teachers who use AI at least weekly estimate they save 5.9 hours a week. Those savings depend on teachers knowing how to review and fix what AI drafts, which is what sustained PD builds.

Sources

  1. Education Week, More schools are providing AI training for teachers. Is it any good? (May 2026, EdWeek Research Center) (opens in a new tab) (checked 2026-11-08)
  2. Education Week, Teacher AI training is rising fast but still has a long way to go (November 2025, EdWeek Research Center) (opens in a new tab) (checked 2026-11-08)
  3. Education Week, More teachers than ever before are trained on AI (April 2025, reporting RAND American School District Panel) (opens in a new tab) (checked 2026-11-08)
  4. Gallup and Walton Family Foundation, teachers receive no formal guidance on AI (May 2026) (opens in a new tab) (checked 2026-11-08)
  5. Gallup and Walton Family Foundation, three in ten teachers use AI weekly, saving six weeks a year (June 2025) (opens in a new tab) (checked 2026-11-08)
  6. Learning Policy Institute, Effective Teacher Professional Development (Darling-Hammond, Hyler and Gardner, June 2017) (opens in a new tab) (checked 2026-11-08)
  7. Kraft, Blazar and Hogan, The effect of teacher coaching on instruction and achievement: a meta-analysis of the causal evidence (Review of Educational Research, 2018) (opens in a new tab) (checked 2026-11-08)
  8. 20 U.S.C. 7801(42), definition of professional development (Elementary and Secondary Education Act, via Cornell Legal Information Institute) (opens in a new tab) (checked 2026-11-08)
  9. U.S. Department of Education guidance on AI and federal grant funds (July 2025, via THE Journal) (opens in a new tab) (checked 2026-11-08)
  10. California Department of Education, Title II, Part A program funding profile (opens in a new tab) (checked 2026-11-08)
Ernie DelgadoFounder and CEO of BeyondK12. 35 years in education technology, founder of Beyond Technology Education in 2001, host of the BeyondK12 podcast.About BeyondK12

Two emails a month for school leaders

Policy changes, funding deadlines, and what is working in real schools, written plainly. Unsubscribe anytime.

By subscribing you agree to receive emails from BeyondK12. See our privacy policy.

See where your school stands on AI, in five minutes

Eight questions. A score out of 24 across Policy, People, Curriculum, Guardrails, and Funding, with the three fixes that move it most. The report is yours whether or not you ever call us.

Get My Free Readiness Score