AI in the Classroom

In this episode we look at a randomized study of students in England that compared three approaches to reading: traditional note-taking, using a large language model (LLM), and combining an LLM with note-taking. Three days later, researchers tested students on retention, comprehension, and free recall.

Topics covered:

• What the study found about AI, note-taking, and reading comprehension
• The gap between what students prefer and what actually improves learning
• The case for having students think and take notes first, then turn to AI
• How student interactions with an LLM could give teachers insight into misconceptions and areas of difficulty
• What the study tells us about AI’s role in reading and learning

Source:

https://www.sciencedirect.com/science/article/pii/S0360131525002829

In this episode we explore what it’s like to go through college with AI always available.

Dan Cogan-Drew sits down with his son Leo, a 21-year-old college senior who began college in 2023, making him part of the first generation of undergraduates whose entire college experience has unfolded alongside widely available generative AI.

Topics covered:

• How a college student actually uses AI
• Using AI as “damage prevention”
• What Leo is comfortable using AI for
• The difference between using AI in humanities and science courses
• Cognitive offloading
• How AI changes the choices students have to make about reading, writing, and studying
• What Leo would tell high school students preparing to enter college in the age of AI

In this episode we explore a consequential week for AI in education, and the growing tension between how quickly technology companies are putting AI into schools and how quickly school systems can decide what responsible adoption should look like.

We examine new AI restrictions from New York City Public Schools and Los Angeles Unified School District, including why two of the nation’s largest school systems are taking such cautious approaches. We also look at the broader environment behind those decisions: Google making Gemini available to students by default, Anthropic expanding Claude into education, and the challenge schools face when powerful new technologies arrive before districts have had the opportunity to evaluate or intentionally adopt them.

Topics covered:

• New York City’s restrictions on student AI use
• LAUSD’s decision to block AI on district devices
• Google making Gemini available to students by default
• Why “opt-out” AI adoption creates challenges for schools
• The importance of keeping educators and administrators in the loop when new AI features launch
• Anthropic’s Claude offerings for teachers and administrators
• Student privacy concerns when educators upload information to AI tools
• The American Psychological Association’s guidance on children, adolescents, and education technology
• The launch of Responsible Technology and Education

Have a story we should be following, or are you experimenting with AI in your classroom, school, or district? Reach us at [email protected].

Sources:

https://apnews.com/article/zohran-mamdani-ai-ban-nyc-schools-647f6a968eea0399521b7934418b1aff

https://www.k12dive.com/news/lausd-restricts-all-students-from-using-ai-tools/829612/

https://www.edweek.org/technology/schools-caught-flat-footed-after-google-makes-gemini-chatbot-available-to-all-students/2026/09

New York City Public Schools has announced a one-year moratorium on student-facing generative AI for grades K–8, a policy affecting roughly 600,000 students and one of the most significant moves yet by a major school district to put limits on AI in the classroom.

In this episode we explore what NYC’s new AI policy gets right, where it may go too far, and the questions educators and district leaders should be asking as schools decide when and how students should use generative AI.

Topics covered:

• What NYC’s one-year K–8 generative AI moratorium actually does• NYC’s new Technology in Schools Coalition• The district’s limited high school AI pilots• Why AI literacy matters• Whether middle schools should have more flexibility to pilot AI tools• What “evidence of instructional impact” should actually mean• The role politics and the technology industry may play in shaping school AI policy• Why NYC’s decision could influence how other major districts approach AI

Sources:

https://www.nyc.gov/mayors-office/news/2026/09/mayor-mamdani-and-chancellor-samuels-put-students-first-with-nat

https://laist.com/brief/news/education/lausd-students-barred-artificial-intelligence-tools

https://www.edsafeai.org/

In this episode we explore the relationship between AI use, belonging, and social connection among young people, and what two recent studies suggest educators should understand as students return to school.

Drawing on research from the Rhythm Project and the California Partners Project, we look at why some young people are turning to AI for more than help with tasks.

Topics covered:
• The connection between students’ sense of belonging and their AI use• Why emotional vulnerability may predict higher-risk AI use• The difference between purposeful AI use and dependency on AI for emotional support• Why students who feel unwelcome in physical spaces may turn to AI chatbots more often• The lack of conversations between young people and adults about AI• Why teachers need to understand students’ out-of-school AI experiences• What schools can do to create stronger opportunities for belonging and human connection

Sources:

https://www.calpartnersproject.org/_files/ugd/2ceb15_fbca2279eb654ca1a91f2c10e756e6ab.pdf

https://drive.google.com/file/d/1oGI3uxseNdfIc2Pms19TsKBQOB7wnJhV/view

In this episode we explore what thoughtful AI use looks like through the eyes of a first-year teacher preparing to enter the classroom.

We speak with Stefanie Stoj, a recent Wesleyan University graduate and Teach For America corps member who is beginning her career teaching seventh-grade ELA in the South Bronx. Stefanie shares how she used tools like ChatGPT and Claude during summer teaching and lesson planning.

Topics covered:

• How a first-year teacher is preparing for her seventh-grade ELA classroom
• Using AI to anticipate student misconceptions
• Using ChatGPT and Claude as lesson-planning tools
• Where AI can save teachers time without replacing professional judgment
• The risk of gradually outsourcing more thinking to AI
• What Stefanie learned about AI use as a college student
• Balancing efficiency with the individual needs of students
• How new teachers can develop their own judgment while working with AI

Source:

https://www.sbecacs.org/

In this episode we explore when AI can actually improve cognition, and when it risks doing the thinking for students.

Building on Dr. Philippa Hardman’s work on cognitive offloading, we examine what separates productive AI use from simply handing cognitive work over to a machine. 

Topics covered:

• When AI use can improve cognition
• The difference between using AI as a collaborator and using it as a substitute for thinking
• Why AI should “critique, never complete”
• The debate over which tasks can safely be offloaded
• How learner motivation changes the risks of cognitive offloading
• Practical design principles for using AI without undermining the learning itself

Source:

https://drphilippahardman.substack.com/p/how-to-design-around-cognitive-offloading

In this episode we explore three stories shaping the conversation around AI in education, and a common thread running through all three: AI alone isn’t enough.

We look at how Khan Academy and Google are evolving Khanmigo beyond the original vision of a student-facing AI tutor. We also examine the idea of AI as an “arrival technology,” technology that entered schools before districts had time to build the policies.

Finally, we dig into new research from OpenAI examining what happens when students use ChatGPT alongside explicit critical-thinking instruction.

Topics covered:
• How Khanmigo is evolving away from a student-facing AI tutor
• Google and Khan Academy’s new AI-supported classroom capabilities
• The challenge districts face when students already use AI outside approved school systems
• OpenAI research on ChatGPT combined with critical-thinking instruction
• What these developments mean for teachers and district leaders

Sources:

https://openai.com/index/what-students-gain-from-chatgpt-critical-thinking-training/

https://www.marketscale.com/industries/education-technology/k-12-ai-spending-is-moving-from-classroom-apps-to-vetting-policy-and-proof

https://blog.google/products-and-platforms/products/education/khan-academy-back-to-school/

In this episode we explore the growing concern around cognitive offloading, what happens when students hand too much of the thinking over to AI.

Drawing on recent research highlighted by Dr. Philippa Hardman, we examine an important distinction: AI itself isn’t necessarily the problem. What matters is when and how it enters the learning process.

Topics covered:

• What cognitive offloading actually means
• How AI-generated feedback can narrow thinking when it arrives too early
• Why deadlines, grades, and output-focused assignments can encourage students to outsource thinking
• How unstructured AI use can increase cognitive load rather than reduce it

Source:

https://drphilippahardman.substack.com/p/how-to-design-around-cognitive-offloading

In this episode we explore how schools can embrace AI without losing sight of what education is actually for.

We sit down with Chris Lehmann, founder and principal of Philadelphia’s Science Leadership Academy (SLA), an inquiry-driven, project-based school built around five core values: inquiry, research, collaboration, presentation, and reflection. Our conversation examines how those values are shaping SLA’s approach to AI—and why decisions about what students can outsource to AI ultimately require schools to be much clearer about what they want students to learn and become.

Topics covered:

• How SLA’s inquiry-driven, project-based, caring model shapes its approach to AI
• SLA’s approach to AI, citation, and plagiarism
• How AI can support students during long-term projects and senior capstones
• What students should—and shouldn’t—outsource to AI
• How technology, algorithms, and information overload are changing what students need from schools
• Why AI forces educators to reconsider the deeper purpose of school
• How schools can preserve student thinking, agency, and human relationships as AI becomes more capable

Source:

https://2027.educon.org/