The Changemaker 6/1/26
Your Teachers Are Already Using AI on IEPs
57% of special education teachers who use AI received no formal training from their district. The question is no longer whether AI is in your IEPs — it's whether you govern it.
AI is already in your IEPs. Do you govern it?
AI is already in your IEPs. Teachers across the country are using generative AI tools to draft IEP goals, write present levels of performance, and generate accommodations language. This is happening without governance, without training, and without district knowledge in most cases.
57% of teachers who reported using AI received no formal AI training from their district. One special education teacher managing 22 students wrote in asking whether there was guidance on AI use in IEP development — her district had none.
The question for SPED directors is not whether AI is being used. It is whether your district will govern it.
That gap is the exposure. The IEP is a legally mandated document, and researchers warn that consumer AI in this process can compromise student privacy, reinforce bias, and weaken the individualized supports that IDEA requires.
Here is the move for this month: Do not treat an AI-in-IEP policy as a future decision. Write the interim version now — which tools are permitted, what student data may never enter a public model, and where a human must stay in the loop. The cost of writing it now is a meeting. The cost of not writing it shows up in a due process hearing.
State funding developments
California
The Governor's budget proposes a 43% increase in special education funding — the largest proposed increase in the state's history ($2.4 billion over the prior year). The proposal still needs to pass the legislature.
Kansas
The state is reimbursing districts at approximately 75.4% of special education costs. The legal requirement is 92% — a benchmark unmet since 2010–2011. Districts operating below adequate reimbursement face compounding deficits in special education staffing and services.
Worth naming plainly: FAPE does not bend to a state budget. A student with a disability is owed an appropriate education whether or not the legislature funds its share. What a shortfall does is force districts to backfill from general funds — a bill that lands on every student in the building.
The AI equity problem in special education
The AI equity problem in special education is not about technology. It is about what districts owe their staff and students.
If a general education teacher has access to an AI writing assistant that saves her two hours a week, and a special education teacher with twice the documentation workload has no training and no tools, that is not a neutral outcome. It is a resource inequity that is compounding an already inequitable workload.
Districts that want to govern AI well in special education need to start with three questions:
- What tools are teachers already using?
- What are the risks specific to IEP documentation — FERPA, accuracy, individualization?
- Who is responsible for setting the standard?
The teachers who are already using AI are not doing something wrong. They are solving a workload problem that their district has not addressed. Meet them there.
Prioritizing general education teachers for AI professional development first leaves the teachers who serve the most complex learners without the tools to keep up. That is not just a professional development gap. It is an equity gap — and it affects students' access to timely, high-quality specially designed instruction.
Get The Changemaker in your inbox every week.
Subscribe · Past Issues