AI in the Classroom

In this episode we explore how teachers are moving beyond restrictive AI policies and building classroom approaches rooted in trust, transparency, and student learning.

Topics covered:

• Why classroom AI policies are shifting from restriction to transparency
• How disclosure can create opportunities for teaching and reflection
• The value of including student voices in policy design
• Why AI use should be understood as a spectrum rather than a yes-or-no question
• How teachers can explicitly teach responsible AI decision-making
• Why classroom policies need to evolve throughout the school year
• How educators can model the transparency they expect from students
• The difference between designing for compliance and designing for integrity

Sources:

https://www.edutopia.org/article/8-classroom-ai-policies-developed-by-teachers/

https://www.edutopia.org/article/ai-transparency-surveys-tool-navigating-student-choices-together

https://wpvip.edutopia.org/wp-content/uploads/2026/07/Class-AI-Agreement_Stacy-Kratochvil.pdf

In this episode we explore why two recent studies—one showing AI weakening students’ math performance and another showing AI-enhanced feedback improving writing—may actually point to the same conclusion: the impact of AI depends on how learning is designed.

Topics covered:

• What research reveals about AI use and declining math performance
• How combining peer and AI feedback can strengthen student writing
• The difference between receiving an answer and exercising judgment
• Why AI exposes assignments built around task completion
• Understanding by Design and the role of backward planning
• How to identify meaningful transfer goals
• Designing writing tasks around audience, purpose, and lived experience
• Creating math assessments that measure genuine problem-solving
• Why stronger assignment design may be more effective than banning AI

Sources:

https://arxiv.org/pdf/2605.21629

https://hechingerreport.org/proof-points-ai-eroding-math-skills/

https://www.sciencedirect.com/science/article/pii/S8755461525000088

In this episode we explore three stories that reveal where AI in education actually stands, not where the industry says it is headed, but what schools are dealing with right now.

Topics covered:

• Why Claude for Teachers raised student-data and FERPA concerns• Why district governance must come before teacher adoption• What the collapse of InBloom can teach today’s AI companies• How AI data centers are affecting laptop and Chromebook prices• Why some schools are pulling back from one-to-one devices• Y Combinator’s vision for AI tutors• The influence of The Diamond Age and “The Primer”

Sources:

https://drjoephillips.substack.com/p/anthropic-launched-claude-for-teachers?r=5qyht

https://www.idahoednews.org/top-news/schools-face-soaring-laptop-costs-as-ai-data-centers-squeeze-supply/

https://www.nytimes.com/2026/03/29/technology/chromebook-remorse-kansas-school-laptops.html 

In this episode we speak with Jim Bentley, a fifth- and sixth-grade teacher in Elk Grove, California, who is entering his 31st year in the classroom. We examine how Jim has begun using AI with students. 

Topics covered:

• The classroom as a laboratory for curiosity, experimentation, and productive mistakes
• How project-based learning shaped Jim’s teaching philosophy
• An AI image-generation project connected to Raymie Nightingale by Kate DiCamillo
• Prompt writing, revision, and healthy skepticism toward AI-generated results
• Media literacy, lateral reading, and fact-checking AI responses
• The difference between using AI as a productivity tool and using it as a learning tool
• What it means for students to do “real work,” not merely “schoolwork”
• Rubric design, teacher judgment, and the risk of overvaluing polished final products

In this episode, part 4 of our series on the productive struggle for teachers, we explore a growing challenge in teacher preparation: how can supervisors tell whether a teacher candidate used AI as a thinking partner, or simply submitted a polished AI-generated lesson plan?

In this episode we explore:

• Why AI can make a teacher candidate’s thinking invisible
• The difference between thoughtful AI-assisted planning and minimal revision
• Why teacher preparation programs should grade decision logs, not just lesson plans
• How supervisors can identify genuine pedagogical reasoning
• Which parts of lesson design should remain productive struggles
• How AI can remove unnecessary friction without replacing professional judgment

In this episode, part 3 of our series on the productive struggle for teachers, we explore how AI is shaping the habits of early-career teachers, and why the choices teachers make in their first five years could influence the rest of their careers.

Topics covered:

• Why early-career teachers may be especially vulnerable to passive AI use• How generic AI-generated materials can affect student engagement• The connection between teacher confidence, burnout, and retention• Why productive struggle matters when learning to teach• The importance of revising and personalizing AI-generated lessons• Why protected reflection time should be part of AI implementation• How early-career teacher support affects equity and the long-term teacher pipeline

In this episode we explore how teachers can use AI during lesson planning without outsourcing the thinking that helps them grow as educators.

We look at why teachers should do the difficult thinking first, how AI can be used to pressure-test a lesson, and why revision and reflection are essential when evaluating AI suggestions. 

Topics covered:

• Why lesson planning is a learning process for teachers• The difference between using AI as a thinking partner and outsourcing the work• Productive struggle and its role in developing teacher expertise• How pedagogical content knowledge is built through planning• Why AI-generated lesson plans require substantial revision• Metacognitive reflection as an alternative to cognitive surrender

Sources:

https://www.nature.com/articles/s41539-025-00311-8#citeas

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6097646

In this episode we explore two major developments shaping AI in K-12 education: Anthropic’s launch of Claude for Teachers and the fragmented policy environment surrounding AI adoption in schools.

Claude for Teachers promises to reduce teacher workload, but the product’s integrations raise important questions about student data, classroom recordings, vendor agreements, and FERPA compliance.

Topics covered:

• What Claude for Teachers offers K–12 educators• How AI could support lesson planning and instructional coaching• The privacy risks of sharing classroom audio and student data• Why vendor integrations require district-level review• The difference between “FERPA-aligned” and FERPA-compliant use• New state laws governing student data and automated decisions

Sources:

https://www.chalkbeat.org/2026/07/14/anthropic-launches-claude-for-teachers-as-ai-companies-battle-for-classrooms/

https://www.youtube.com/watch?v=V-OOEC5RNaQ

In this episode we explore what happens when teachers use AI to save time, and whether those shortcuts can come at a cost for students.

Drawing on a statewide survey of more than 13,000 Georgia teachers, we examine a troubling tension: Teachers often believe AI improves their work, while emerging evidence suggests that certain patterns of AI use may reduce student motivation, confidence, and, in some classrooms, academic performance.

Topics covered:

• Why teacher AI adoption is accelerating• The gap between perceived productivity and student outcomes• Findings from Georgia’s statewide teacher survey• What a randomized controlled trial found about student motivation and confidence• Cognitive offloading and the “skill substitution” problem• How generic AI-generated materials can weaken a teacher’s personal voice• The difference between using AI as a shortcut and using it as a thinking partner

Sources:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7007339

https://www.the74million.org/article/more-than-half-of-georgia-teachers-now-use-artificial-intelligence-to-prepare-for-class/

https://www.anthropic.com/news/claude-for-teachers

In this episode we explore what happens when teachers use AI to prepare lessons, assessments, discussion questions, and other classroom materials, and whether working faster actually leads to better learning.

As OpenAI, Google, Microsoft, and Anthropic race to integrate AI into educators’ daily workflows, the promise is straightforward: save teachers time. But emerging research suggests that when teachers rely too heavily on AI-generated materials, students may find classes less interesting, feel less motivated, and, in some cases, perform worse.

Topics covered:

• The rapid growth of AI-assisted lesson preparation• The difference between faster content and better instruction• Research linking passive AI use to lower student motivation and engagement• Why students notice when classroom materials feel generic or impersonal• The launch of Claude for Education and integrations with teacher productivity tools• How time pressure and limited training influence teacher AI use• Why revising and personalizing AI-generated materials matters

Sources:

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7007339

https://www.the74million.org/article/more-than-half-of-georgia-teachers-now-use-artificial-intelligence-to-prepare-for-class/

https://www.anthropic.com/news/claude-for-teachers