AI for Professional Development — Training Teachers on New Technology

Professional development for technology adoption has a dismal track record. Research from the Bill and Melinda Gates Foundation (2014) found that only 29% of teachers felt that PD activities they attended were useful — and technology-focused PD scored even lower. The dominant model — "sit in a room while someone clicks through slides about a tool you didn't choose" — fails because it violates nearly every principle of adult learning. It's passive, decontextualized, one-size-fits-all, and disconnected from teachers' actual classroom challenges.

AI PD faces these same challenges, plus additional ones. AI is not a single tool with a fixed interface — it's a capability that changes monthly. The skills teachers need aren't button-clicking procedures but judgment: knowing when AI helps, when it doesn't, how to evaluate its output, and how to integrate it into existing instructional practice. This requires fundamentally different PD design than traditional technology training.

The research on effective PD is clear and consistent across multiple meta-analyses (Desimone, 2009; Darling-Hammond et al., 2017). Effective PD is: (1) sustained over time, not one-shot; (2) content-focused, not tool-focused; (3) practice-based, with time to try and reflect; (4) collaborative, with peer interaction; and (5) supported by coaching, not just workshops. This article translates those principles into specific designs for AI-focused professional development.

---

Why Traditional Tech PD Fails for AI

---

A Tiered PD Framework

Tier 1: Awareness (All Staff — 2-3 Hours)

Goal: Every staff member understands what AI is, what the district's policy is, and how it might affect their role.

Format: Full-staff session (in-person or virtual) with interactive components

Session Design:

SESSION: AI Awareness for Educators (2.5 hours)

PART 1: What AI Actually Is (30 minutes)
- Live demonstration: Show AI generating a lesson plan,
  differentiating a text, and creating an assessment
- Show a failure: Ask AI to generate something it does
  poorly (factual errors, hallucinations, lack of
  nuance) — this builds critical judgment
- Interactive: Teachers try one prompt on their phones/
  laptops right now. Give them exactly this prompt:
  "Create a 5-question quiz on [your subject/topic]
  for [your grade level]."
  Then ask: "How many questions were good? How many
  needed editing? What would you change?"

PART 2: District Policy Overview (20 minutes)
- What's approved, what's not
- Student data rules (what you can/can't enter into AI)
- Academic integrity expectations
- Where to get help and ask questions

PART 3: Use Case Gallery Walk (40 minutes)
- Set up 6-8 stations around the room, each showing a
  specific use case with before/after examples:
  Station 1: Lesson plan generation
  Station 2: Differentiated materials
  Station 3: Assessment creation
  Station 4: Parent communication
  Station 5: IEP goal writing
  Station 6: Rubric creation
  Station 7: Feedback on student writing
  Station 8: Professional email composition
- Teachers rotate, try each, and rate: "Would this
  save me time? Would I use this?" (sticky dot voting)

PART 4: Q&A and Concerns (30 minutes)
- Anonymous question submission (via Google Form or
  index cards)
- Address concerns honestly — don't dismiss fears
- Close with: "Your homework is to try ONE use case
  from the gallery walk this week. Just one."

MATERIALS NEEDED: Laptops/phones, approved AI tool access,
gallery walk station materials (printed examples at each
station), sticky dots for voting, anonymous Q&A form

Tier 2: Foundation (Interested Teachers — 8-12 Hours Over 4-6 Weeks)

Goal: Teachers develop basic competence with AI tools for their specific subject and grade level.

Format: Workshop series (4 sessions × 2-3 hours) + between-session practice

Session Design:

SESSION 1: Effective Prompting (2.5 hours)
- The anatomy of a good prompt: context + task +
  constraints + format
- Practice: Teachers write prompts for their own content
- Peer review: Partners evaluate each other's AI output
- Homework: Generate one week of warm-up activities
  using AI. Grade the output: what's usable as-is,
  what needs editing, what should be discarded?

SESSION 2: Differentiation with AI (2.5 hours)
- Generating tiered materials from a single lesson
- Scaffolding and accommodation embedding
- Practice: Each teacher generates a differentiated
  lesson for their hardest-to-differentiate unit
- Homework: Use the differentiated materials in class.
  Bring back: What worked? What didn't?

SESSION 3: Assessment and Feedback (2.5 hours)
- Generating assessment items (quiz, test, performance
  task) with AI
- Using AI to generate rubrics
- Quality control: How to catch bad questions,
  factual errors, and bias
- Practice: Create an assessment for an upcoming unit
- Homework: Administer the AI-assisted assessment.
  Compare quality to your usual assessments.

SESSION 4: Integration and Sustainability (2.5 hours)
- Building AI into your weekly workflow
- Time-saving strategies: What to AI, what not to AI
- Sharing successes and failures from homework
  assignments
- Creating a personal AI toolkit: Your top 5 prompts
  for your specific teaching context
- Commitment: "What will you continue to do with AI
  after this series ends?"

Tier 3: Advanced Application (Active Users — Ongoing)

Goal: Teachers who are already using AI effectively deepen and refine their practice.

Format: PLC-embedded, monthly sessions + peer coaching

Design Principles:

• Teacher-driven topics — they identify what they want to learn next
• Peer sharing — teachers demonstrate their most effective AI practices
• Problem-solving — bring a challenge, the group helps solve it
• Innovation — experimenting with new applications and reporting back
• Quality focus — how to improve the quality of AI output, not just the quantity

Tier 4: Leadership (AI Coaches — Year-Long)

Goal: Build internal capacity to sustain AI PD beyond this year.

Cohort Model:

• 5-10 teachers identified as potential AI coaches
• Monthly cohort meetings focused on coaching skills (not just AI skills)
• Each coach supports 8-10 colleagues informally
• Coaches observe and provide feedback on AI-integrated lessons
• Coaches contribute to policy refinement based on classroom experience

---

Adult Learning Principles Applied to AI PD

---

Coaching Models for AI

One-shot workshops don't change practice. Coaching does. Joyce and Showers (2002) found that only 5% of teachers transfer a new skill to classroom practice from a workshop alone, but 95% transfer when coaching is added.

AI COACHING CYCLE (Adapted from Jim Knight's
Instructional Coaching Model):

STEP 1: IDENTIFY (1 conversation, 15 minutes)
Coach and teacher identify a specific instructional
challenge that AI might address.
"What takes you the most time each week?"
"Where do you wish you had better materials?"
"What's your most challenging differentiation need?"

STEP 2: LEARN (1 session, 30-45 minutes)
Coach demonstrates the AI application targeting the
teacher's identified challenge. Teacher observes.
"Here's what I'd type into the AI for your Grade 4
fractions unit. Watch what it generates. Let's
evaluate it together."

STEP 3: IMPROVE (1-2 weeks of practice)
Teacher practices independently. Coach checks in
briefly (email, 5-minute hallway conversation).
"How did the AI-generated materials work? What did
you need to modify?"

STEP 4: REFLECT (1 conversation, 15 minutes)
Teacher and coach discuss what worked, what didn't,
and what to try next.
"Would you use AI for this purpose again? What would
you do differently?"

STEP 5: SUSTAIN (ongoing, as needed)
Coach remains available. Teacher is now independently
using AI for this application and ready to tackle a
new one.

CYCLE DURATION: 3-4 weeks per coaching cycle
COACHING LOAD: One coach can sustain 8-10 teachers
simultaneously across different coaching cycle stages.

---

Addressing PD Resistance

---

Measuring PD Effectiveness

---

Key Takeaways

• Traditional tech PD fails for AI because AI isn't a traditional technology. It's a capability that requires judgment, not just skills. PD must teach teachers to think with AI, not just click buttons. Design PD around problems teachers actually face, using their actual curriculum content.
• Tiered PD respects the readiness spectrum. Not all teachers need the same training at the same time. Awareness for all, foundation for the willing, advanced application for the active, leadership development for the coaches. EduGenius supports this spectrum — teachers at any comfort level can generate educational content by describing what they need in natural language.
• Coaching, not workshops, changes practice. Joyce and Showers (2002) demonstrated that workshops alone produce only 5% skill transfer; coaching raises this to 95%. Invest in internal AI coaching capacity. One coach supporting 8-10 teachers will produce more adoption than five full-day workshops for the whole staff.
• Make PD opt-in, not mandated, in Year 1. Forced AI training breeds resentment. Build a critical mass of enthusiastic early adopters through voluntary, high-quality PD. Their success stories will recruit the majority more effectively than any mandate.
• Measure what matters: are teachers actually using AI, and is instruction improving? Post-workshop satisfaction surveys are insufficient. Track actual usage, quality of AI-assisted materials, and ultimately student learning outcomes.

See AI for School Leaders — A Strategic Guide to Transforming Education Administration for the strategic leadership framework. See Data-Driven Decision Making in Schools with AI Analytics for using AI in school improvement planning. See Best AI Content Generation Tools for Educators — Head-to-Head Comparison for tool evaluation to support PD planning.

---

Frequently Asked Questions

How many PD hours should we allocate for AI in the first year?

Research on effective PD suggests a minimum of 20 hours of sustained, content-focused learning to produce measureable changes in teacher practice (Desimone & Pak, 2017). For AI, this translates to: 2-3 hours of awareness PD for all staff, plus 10-12 hours of foundational training for active participants, plus ongoing monthly PLC time for reinforcement. Don't front-load all hours in August — spread them across the year with practice time between sessions.

Should AI PD be mandatory or voluntary?

Voluntary for Tier 2+ in Year 1. Mandatory awareness (Tier 1 — understanding policy and basics) is reasonable because it affects the whole school. But forcing resistant teachers into hands-on AI workshops produces compliance without competence and can generate backlash that sets adoption back. Make Tier 2+ voluntary, time-convenient, and clearly useful — then let results speak for themselves.

What if our teachers have wildly different technology comfort levels?

This is normal and expected. The tiered model addresses it directly: Tier 1 gives everyone a baseline. Teachers who are already tech-comfortable will naturally move to Tier 2. Teachers who are tech-anxious may need additional one-on-one coaching support (Tier 4 coaches) before they're ready for group workshop formats. Never assume that a teacher who's uncomfortable with technology can't benefit from AI — some of the most effective AI users are teachers who are weak technologists but strong pedagogues, because they know exactly what good instruction looks like and can evaluate AI output against that standard.

How do I sustain AI PD after the initial rollout year?

Three mechanisms: (1) Internal AI coaches (Tier 4) who continue supporting teachers after external PD funding ends. (2) AI integration into existing PLC and department meeting structures — not additional meetings, but AI as a standing agenda item in meetings that already happen. (3) Peer sharing structures where teachers who develop effective practices share them informally (a shared drive of "best prompts," a monthly 15-minute show-and-tell, a Slack channel for AI tips).

---

Next Steps

• AI for School Leaders — A Strategic Guide to Transforming Education Administration
• Data-Driven Decision Making in Schools with AI Analytics
• Budgeting for AI in Education — ROI, Costs, and Funding Sources
• AI Policy Development for Schools and Districts
• Best AI Content Generation Tools for Educators — Head-to-Head Comparison