Educator Preparation · AI Essentials
AI in Teacher Preparation: Why Colleges of Education Need Faculty Adoption and Student Readiness
Nearly every conversation about AI in teacher preparation is about the candidates. Almost none of it is about the professors preparing them. That asymmetry is quietly becoming the biggest credibility risk in educator preparation.
AI in teacher preparation has become a one-sided conversation. Walk into almost any College of Education right now and you will find some version of the same initiative: a new module on AI literacy for teacher candidates, a revised syllabus statement, maybe a task force on academic integrity. The candidates are the focus. They are the future teachers, after all, and they are walking into K-12 classrooms where AI tools are already in daily use.
But ask a quieter question, the one deans tend to raise only behind closed doors: how many of our own faculty could confidently model responsible AI use in front of those candidates tomorrow?
For most programs, the honest answer is uncomfortable. And that gap between what we expect candidates to learn and what faculty are equipped to demonstrate is the part of AI in teacher preparation nobody is writing about.
The asymmetry problem in AI in teacher preparation
Teacher preparation has always operated on a simple premise: faculty model the practice, candidates learn the practice, and K-12 students eventually benefit from the practice. Modeling is not a nice-to-have in teacher prep. It is the pedagogy. We teach future teachers how to teach by showing them.
AI has broken that chain in a way no previous technology quite did. When interactive whiteboards or learning management systems arrived, faculty could adopt them gradually, a semester or two ahead of expectations. Generative AI arrived in candidates' pockets first. Many teacher candidates today have more hands-on hours with AI tools than the professors evaluating their lesson plans.
The result is a credibility inversion that educator preparation has rarely faced: the learners are ahead of the modelers. And every program that responds by adding another candidate-facing module, without touching faculty capability, widens the gap it is trying to close.
What faculty AI adoption actually protects
It is tempting to file faculty AI development under professional courtesy, something to offer when budgets allow. That misreads what is at stake. Faculty adoption protects three things a College of Education cannot function without.
1. Instructional credibility
Candidates are being told AI fluency is a professional expectation. If the program delivering that message cannot demonstrate the expectation internally, candidates draw the obvious conclusion: this is a compliance topic, not a practice. The fastest way to make AI literacy feel optional is to have it taught by people who visibly opted out.
2. Curriculum quality
Faculty who have never used AI to differentiate feedback, draft an IEP-aligned accommodation, or pressure-test a lesson plan cannot design authentic assignments about those tasks. They can only design assignments about AI in the abstract. The difference shows up immediately in candidate work: reflection essays about AI instead of demonstrated judgment with AI.
3. Accreditation posture
This one is not hypothetical. CAEP Standard 6 already expects educator preparation providers to offer professional development for faculty, including training in the use of technology, and infrastructure that supports faculty and candidate use of technology in instruction. Faculty AI adoption is not a stretch of that language. It is the most current reading of it. When the self-study asks how the program supports faculty and prepares candidates for technology-rich classrooms, "we added a module" is a thinner answer than "our faculty are trained, our candidates are assessed, and here is the documentation for both."
Why candidate-only AI in teacher preparation falls short
None of this argues for shifting attention away from candidates. Their readiness is the product a College of Education ships. The argument is that candidate-only approaches to AI in teacher preparation have a structural ceiling, and programs keep hitting it.
Faculty-blind approach
Train candidates only
- AI content lives in one course or module, owned by one enthusiastic instructor
- Candidates receive mixed signals as expectations change from professor to professor
- Assignments test knowledge about AI, not judgment with AI
- Momentum leaves when the champion faculty member does
Dual-track approach
Train faculty and candidates together
- Consistent expectations and language across the whole program of study
- Faculty model live AI use, then candidates practice it in methods courses
- Assessment measures demonstrated, human-in-the-loop judgment
- Capability lives in the institution, not one person
Programs that train only candidates also create a strange durability problem. A candidate cohort graduates every year and takes its training with it. Faculty stay. Faculty capability is the only AI investment a College of Education makes that compounds year over year, cohort after cohort. It is the difference between buying readiness annually and owning it.
A 4-step framework for AI in teacher preparation
The good news for deans: a dual-track approach to AI in teacher preparation is not two separate initiatives with two separate budgets. Done well, the tracks reinforce each other. Here is the four-step sequence we see work.
Benchmark both populations first
Run the same baseline AI readiness assessment across faculty and candidates. Leaders are consistently surprised by the results in both directions, and the data reframes faculty training from remediation to shared professional practice. It also gives you the pre-measure your accreditation evidence will need later.
Give faculty a head start, not a hurdle
Faculty development should run one step ahead of candidate expectations, framed around their real workflows: feedback at scale, differentiated materials, assessment design, advising. When training saves faculty time in week one, adoption stops being a mandate and starts being self-sustaining.
Move AI from a module into methods
Once faculty are equipped, AI practice belongs inside existing methods and clinical coursework, where candidates already learn to plan, assess, and differentiate. Simulation-based practice with real tools, followed by reflection on judgment, produces portfolio artifacts a hiring principal can actually evaluate.
Document everything as evidence
Pre- and post-assessment data, faculty completion records, candidate artifacts, and a written AI governance stance together form an evidence trail that serves accreditation self-studies, state program approval, and grant reporting simultaneously. The same work, counted three times.
The question worth asking this semester
Colleges of Education sit in an unusual position in the AI transition, and it makes AI in teacher preparation different from AI adoption anywhere else on campus. They are not just adopting a technology for their own institution. Every faculty member they equip shapes dozens of candidates, and every candidate shapes thousands of K-12 students over a career. The multiplier runs through the faculty, or it does not run at all.
So the question for program leaders is not whether to add more AI content for candidates. Most programs are already doing that. The question is simpler and harder: if an accreditor, a district partner, or a candidate asked your faculty to demonstrate the AI practices your program teaches, how would that go?
If the answer gives you pause, that pause is the strategy talking. AI in teacher preparation on one track is not readiness.
Benchmark your faculty and your candidates
AI Essentials partners with Colleges of Education to assess AI readiness across both populations, deliver simulation-based training aligned to educator preparation standards, and produce the documentation your program approval and accreditation work requires.
Talk to our team