AI in the Classroom

When is an educational model actually ready to scale?

Alpha School has drawn attention for its promise of compressing traditional academics into a two-hour, AI-powered learning block, along with claims of dramatic student growth. But beneath the headlines is a much bigger question about the role we expect software to play in teaching and learning.

Topics covered include:

• Why Alpha School’s model is attracting both attention and skepticism
• What evidence should exist before an educational innovation scales
• The enduring influence of Benjamin Bloom’s “2 sigma” tutoring research
• What effective human tutors do that software may struggle to replicate
• What recent AI tutoring research suggests about student engagement
• The idea of the “closed loop fallacy”
  

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Have a story we should cover? Reach us at [email protected]

Sources:

https://michaelbhorn.substack.com/p/could-alpha-school-disrupt-k12-education

https://michaelbhorn.substack.com/p/if-you-dont-like-alpha-its-probably

In this episode we explore why I stopped using AI as a thinking partner in developing this podcast.

For more than 100 episodes, AI was part of the show’s development process. But after a conversation with a high school English teacher, and some reflection on how the workflow itself had changed, I began to question whether the tool was still helping me think more clearly.

Topics covered include:

• How an AI thought partner can gradually shift your thinking
• Why efficiency isn’t always the same as better thinking
• Where AI still feels useful in the podcast workflow
• The strange experience of hearing an AI-generated version of your own voice
• What this experiment might suggest for teachers reconsidering their own use of AI

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Have a story we should cover? Reach us at [email protected]

Source:

https://pz.harvard.edu/resources/i-used-think-now-i-think

In this episode we explore what happens when concerns about AI cheating begin to reshape the relationship between students and educators.

We also look at the growing divide between policies that signal trust and those that signal mistrust.

Topics covered include:

• Why AI-related cheating concerns are becoming a trust issue• What students say about being falsely accused of using AI• The unintended consequences of AI detection• What recent school-system AI bans may actually be signaling• How teachers can rethink assignments for an AI-enabled world

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Have a story we should cover? Reach us at [email protected]

Sources:

https://www.edweek.org/technology/how-ai-suspicions-are-undermining-student-teacher-relationships/2026/09

https://digitalthriving.gse.harvard.edu/an-ai-policy-isnt-a-playbook/

https://open.spotify.com/episode/28y6bsFpSmPQlcwoTI0yek

In this episode we explore three stories that reveal just how differently schools are beginning to respond to AI, from redesigning major courses and assessments, to confronting parent concerns about AI tutors, to using AI behind the scenes to rethink school transportation.

Topics covered include:

• The College Board redesigning AP Computer Science Principles to incorporate generative AI, machine learning, data ethics, and a new AI-focused performance task
• Why AP Seminar is changing its approach to writing and AI
• The growing mismatch between school assignments and the AI-enabled world students are entering
• SFUSD’s use of the AI-powered reading tutor Amira
• How one New York school district is using AI to optimize bus routes in real time
• Why some of the least controversial uses of AI in schools may be happening far outside the classroom

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Have a story we should cover? Reach us at [email protected]

Sources:

https://newsroom.collegeboard.org/new-college-board-research-ap-teachers-push-guardrails-and-support-genai-reshapes-classroom 

https://sfstandard.com/pacific-standard-time/2026/09/25/pst-sf-amira-ai-in-schools/

https://www.govtech.com/education/k-12/frontier-schools-n-y-turn-to-ai-for-bus-routing

In this episode we explore what it looks like when a school leader decides to take AI seriously, without letting the technology dictate what good teaching and learning should become.

We sit down with Vinnie Murray, Principal of Dominican High School for a conversation about Dominican’s evolving approach to AI.

We also explore the tension Vinnie is seeing in his own students: many are skeptical of AI even as they continue to use it.

Topics covered include:

• How Vinnie moved from experimenting with ChatGPT to leading a school-wide AI strategy
• Why teacher buy-in may matter more than getting everyone to the same level of enthusiasm
• What student surveys at Dominican reveal about changing attitudes toward AI
• How the conversation is shifting from “AI and cheating” to assessment and instructional design
• How teachers can make student thinking more visible when work happens outside the classroom
• How the school’s values are shaping decisions about when students should use AI

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

In this episode we explore what educators are actually seeing when AI-powered tools move from demos and headlines into real classrooms.

Drawing on Instruction Partners’ recent “AI in Action” learning tour, we examine what seems to be working, what clearly isn’t, and why the difference may have less to do with whether a product uses AI than with how that AI is designed and deployed.

Topics covered include:

• What remains central as AI tools become more capable• What researchers observed about AI’s ability to take the pedagogical lead• Why some AI tools may be worse than paper and pencil• The problem with general-purpose chatbots• Why browser lockdowns and academic-integrity policies may only be the beginning• What happens when AI arrives at school through students’ everyday devices• The design principle behind some of the strongest AI-supported learning experiences observed on the tour

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Source:

https://ailearningtour.instructionpartners.org/insights/overall-takeaways

In this episode we explore how to tell whether educational technology is actually helping students learn.

We sit down with Dr. Nicole Barnes, a learning scientist at the American Psychological Association, to unpack a new APA report on children, adolescents, and educational technology.

Topics covered include:

• Why “engagement” can be a misleading measure of learning
• The difference between clicking, completing, and understanding
• What generative AI may change about productive struggle in the classroom
• How teachers can sequence AI use so student thinking still comes first
• What age-appropriate AI guardrails might look like
• What learning science can contribute to decisions schools are already making about AI

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Source:

https://www.apa.org/pubs/reports/childrens-adolescents-learning-educational-technology.pdf

In this episode we explore what AI tutoring looks like when the goal is to recreate some of the things effective human tutors actually do.

Using the AI tool Ello as a case study, we examine a framework drawn from high-dosage human tutoring: understanding what a student partially knows, recognizing when it’s time to shift gears, and building enough of a relationship that the learner wants to keep going.

Topics covered:

• What separates an AI tutor from an AI “teacher”
• Three moves that effective human tutors make
• How AI can listen for a child’s partial understanding
• Why responsiveness and latency matter for student engagement
• The role of emotion and relationship in tutoring
• What happens when children become attached to an AI character
• The guardrails developers are building around those relationships
• The growing tension between school restrictions and AI adoption at home
  

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Sources:

https://humanistxyz.substack.com/p/what-it-takes-to-build-an-ai-teacher

https://www.ello.com/

In this episode we explore three stories around how schools can make smart decisions about AI while the evidence is still catching up.

We look at a Stanford review of hundreds of studies on AI in K–12 education and what it reveals about learning. We also examine how school districts are moving beyond broad “AI policies” toward rules for specific classroom uses.

Finally, we explore a new open infrastructure project designed to give educators and EdTech developers better access to academic standards, learning progressions, and tools for evaluating AI-generated educational content.

Topics covered:
• Why better performance with AI doesn’t necessarily mean better learning• What happens when students have to perform without the AI scaffold• Where AI may have more potential• How districts are approaching AI differently across specific classroom use cases• Why AI policies might need to be treated as testable hypotheses rather than permanent rules• What educators should know about new open infrastructure for AI and EdTech

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Sources:

https://www.nctq.org/research-insights/ai-can-help-students-and-teachers-but-does-it-improve-learning/

https://www.edsurge.com/news/learning-commons-launches-open-ai-platform-for-k-12-edtech

https://www.wbur.org/news/2026/09/22/massachusetts-schools-ai-use

In this episode we explore the question: What does “authentic” student work mean when AI is increasingly part of authentic work in the real world?

Prompted by a report about British members of Parliament using AI to help write speeches, we explore a tension at the heart of assessment design. If professional coders use AI to code and politicians use AI to write speeches, should students be expected to do those things without it?

In this episode we explore:

• How AI is exposing a long-standing relevance problem in school assignments
• Why “authentic assessment” may be getting harder to define
• The tension between how students are expected to learn and how professionals actually work
• What coding, political speeches, and traditional English assignments have in common
• What educators may need to preserve as more human work becomes AI-assisted
  

AI in the Classroom Daily helps educators make sense of AI without the hype.

Hosted by Dan Cogan-Drew, co-founder and Chief Academic Officer at Newsela.

Sources:

https://ncee.org/scaling-the-school-engagement-cliff/

https://www.hightechhigh.org/

https://www.economist.com/britain/2026/09/23/ai-written-speeches-are-taking-over-politics

https://www.nytimes.com/2026/09/23/opinion/ezra-klein-podcast-jensen-huang.html

https://gptzero.me/ai-vision