The American Association of Colleges of Teacher Education released a new framework today calling for university-based teacher education programs to build a comprehensive approach to preparing teachers for AI's role in K-12 classrooms. The association, which represents more than 500 public and private college and university programs, counts on those programs to prepare more than 90% of new educators, according to federal estimates.
The framework also suggests that aspiring teachers who oppose AI on ethical grounds should be allowed to avoid using it and still receive their teaching credentials - a proposal that could be tricky for states to implement.
Developed by a task force of members and outside experts, the framework lays out the technical skills, pedagogical knowledge, and ethical judgment teachers need to use and teach AI well. It does not include specific curriculum examples or standards. AACTE considers it a starting point for deeper conversations among higher education institutions.
What the framework asks of teacher training
The framework says preservice teachers need explicit training in four areas:
- Ethics and policy, such as AI literacy and equitable access.
- Practice and implementation, including evaluating student work produced with AI tools.
- Cognitive issues in the technology - biases and hallucinations - and in how AI affects people, such as its impact on students' critical thinking skills.
- Professional expertise, including how teachers use their own judgment and build teacher-student relationships.
AI's implications touch many fundamentals of teaching candidates learn, including planning, assessment, differentiation, instructional choices, and professional ethics, said Anne Tapp Jaksa, a teacher education professor and senior research director of the University of Michigan's Center for Digital Curricula who helped develop the framework.
"AI forces us to reconsider what counts as evidence of learning," Tapp Jaksa said. "If a candidate can use AI to generate a polished lesson plan or reflection or an assessment, the important question becomes whether that product actually demonstrates the candidate's own professional reasoning."
Teacher training on AI to date has typically come from individual districts or professional organizations - often developed with AI industry groups and focused on learning specific platforms or prompts, said Peña Bedesem, co-chair of the committee that developed the framework and an associate professor of special education at Kent State University in Ohio.
It's important that teacher education on AI "not become just a computer science course," Bedesem said. "That's not going to be efficient, and it's not going to be effective."
Rethinking what teaching looks like
If teacher education has traditionally been designed around teaching to the average student, "there is no such thing as average when you think about what we can do today with AI to design and structure what learning looks like in different ways," said Carole Basile, dean of the Mary Lou Fulton College for Teaching and Learning Innovation at Arizona State University. "So we're beginning to ask a different set of questions around what teaching could look like."
Basile said ASU's teacher college has overhauled its curriculum to focus on how instruction could evolve in the wake of AI. Teacher preparation, she said, must move from training teachers to use individual tools to helping teachers make their own tools and understand how AI can support, rather than replace, their professional judgment and relationships with students.
For educators looking to build these skills now, AI for Teachers Courses offer practical training on classroom integration. Broader resources on AI for Education cover tools and curriculum approaches for K-12 settings.
The right to refuse AI
AACTE also calls for teacher training and licensure programs to give teachers the right to refuse to use the technology altogether.
"AI is becoming increasingly difficult to avoid, but there may be a legitimate privacy, accessibility, cultural, pedagogical, or ethical reason not to use a particular system," said Tapp Jaksa. Teachers should not be prevented from attaining a teaching license because they choose not to use AI.
But not using AI may be harder once teachers have entered the classroom. Much of the more than $1.7 billion in Title II teacher-training grants now prioritizes preparing educators to use AI in their classrooms, per an April 2025 executive order on AI-based teacher training signed by President Donald Trump.
Why this matters for education professionals
For teacher educators and administrators, the framework signals that AI training can no longer be an add-on or an afterthought. The practical question is how to weave AI literacy into existing coursework without turning it into a standalone computer science class - while still respecting faculty and students who want no part of it. Programs that start these conversations now, rather than waiting for state mandates, will have more room to shape their own approach.
Your membership also unlocks:
