AI in the Classroom - Daily

In this episode we explore what happens when AI changes the basic conditions of student writing, and why some teachers are bringing pencils, notebooks, and in-class writing back into the center of English instruction.Drawing on a recent New York Times piece about AI and student writing, we look at the growing move away from take-home essays and toward handwritten, in-class work. The real question may not be paper or AI. It may be something more.Topics covered:Why teachers are rethinking take-home writing assignments in the age of AIWhat student AI use reveals about homework, drafting, and critical thinkingHow handwritten assessments can unintentionally measure the wrong thingsThe difference between drafting, revising, and composing under time pressureWhy AI should support the writing process without replacing student authorshipWhat district leaders should consider before making “back to paper” policiesA practical writing workflow that combines notebooks, rubrics, AI feedback, and student reflectionSources:https://www.nytimes.com/2026/04/30/us/ai-students-cheating-homework-classrooms.html

In this episode we explore a new Stanford study showing that AI writing feedback tools can respond differently to the same student work depending on demographic cues like race, gender, motivation level, English learner status, or learning disability status.We break down why this matters for classrooms and districts, especially as schools adopt AI feedback tools that promise personalization. The central question: when does “personalized” feedback actually support students, and when does it quietly reinforce stereotypes?Topics covered:What Stanford researchers found when they tested AI feedback on middle school writingWhy some students received more praise while others received more substantive revision guidanceThe difference between encouragement and feedback that actually helps students growWhy demographic data can both help districts monitor equity and introduce new forms of biasHow this differs from culturally responsive teachingWhat teachers can do to help students reflect critically on AI-generated feedbackWhat instructional coaches can test inside their own schoolsWhat district leaders should ask vendors before approving AI writing toolsSources:https://arxiv.org/pdf/2603.12471https://hechingerreport.org/proof-points-ai-bias-feedback/

In this episode we explore the promise, and the limits, of “closed loops” in AI-enhanced classrooms.We talk about whether AI can really give teachers and school systems a complete picture of a student’s learning journey: where they are, what they need next, and how best to support them.Topics covered:What “closed loop” means in AI-supported teaching and learningThe appeal of real-time insight into student progressWhy AI may work well for highly motivated students but miss many othersThe danger of mistaking niche AI learning models for universal solutionsWhat technology can measure — and what only teachers can understandWhy student motivation, context, and meaning-making still matterWhat district leaders should watch for when evaluating AI-driven personalizationWhy classroom relationships remain central, even in data-rich environments

In this episode we explore what Khanmigo’s uneven rollout reveals about AI tutoring, student motivation, and the limits of chatbot-based learning tools.When students do not use an AI tutor, is that a student motivation problem, a product design problem, or a teaching design problem?Topics covered:What Khanmigo’s rollout tells us about AI tutoring in real classroomsWhy low student usage matters more than edtech hypeThe difference between a teaching problem and an interface problemWhy students may not know how to ask AI tools for helpHow chatbot interfaces can increase cognitive loadWhat teachers already understand about motivation, confidence, and relationshipsWhy Socratic AI tools may not work for students who are already confusedWhat district leaders should ask vendors about AI usability and evidenceSources: https://danmeyer.substack.com/p/rip-khanmigo-and-edtech-industryhttps://www.chalkbeat.org/2026/04/09/sal-khan-reflects-on-ai-in-schools-and-khanmigo/https://www.oneusefulthing.org/p/claude-dispatch-and-the-power-of

In this episode we explore what happens when school districts begin using “vibe coding” to build their own software tools with generative AI.We look at Peninsula School District in Washington State, where administrators are using AI coding tools to create internal apps and potentially save hundreds of thousands of dollars in EdTech contracts. We examine why this is promising, why it is risky, and what district leaders need to consider before AI-built tools touch student or staff data.Topics covered:What “vibe coding” means in a K–12 school district contextHow Peninsula School District is using AI to build internal software toolsWhy districts may see AI-built apps as a way to reduce EdTech costsThe cybersecurity risks of AI-generated codeThe risks of in-house tools that outlive the people who built themQuestions district leaders should ask before deploying AI-built applicationsWhat teachers should ask when a district introduces an internally built AI toolSource: https://www.k12dive.com/news/vibe-coding-helped-a-washington-district-save-250k-in-ed-tech-costs/816993/

In this first Teacher Tuesday episode of AI in the Classroom – Daily, we explore what AI looks like from inside a real elementary classroom.Dan Cogan-Drew talks with educator Diana Betancourt of Neabsco Elementary School in Prince William County, Virginia, about how AI-supported writing tools are helping her students—especially multilingual learners and students with disabilities—build confidence, organize their ideas, and take greater ownership of the writing process.Rather than treating AI as a shortcut or replacement for teaching, this conversation looks at how carefully designed AI feedback can support student thinking, reduce teacher bottlenecks, and create more opportunities for one-on-one instruction.Topics covered:Why Diana became an educator and how her work with multilingual learners shapes her teachingWhat writing challenges look like for her studentsHow AI feedback can help students move from ideas to organized writingWhy confidence matters so much for English learners and students with disabilitiesHow AI can support metacognition without giving students the answerWhat teachers should, and should not, expect AI to do in the classroomWhy AI should amplify teacher judgment, not replace itWhat district leaders can learn from teachers experimenting thoughtfully with AISource: https://wtop.com/prince-william-county/2026/03/how-a-prince-william-co-teacher-is-using-ai-to-offer-students-immediate-feedback/

In this episode we explore whether AI translation tools are making language learning obsolete, and why the answer is more complicated than it first appears.Using a recent essay from The Conversation as a starting point, we look at the difference between using AI for quick transactional communication and using language to build identity, relationships, and cultural understanding.Topics covered:AI translation tools and the future of language learningThe difference between transactional communication and identity-based communicationWhy learning another language is about more than exchanging informationWhat this means for world language classroomsHow teachers can design assignments that students are less likely to offload to AIWhy personal investment, identity, and ownership matter in AI-era learningHow educators can decide which parts of learning should stay humanSource:https://theconversation.com/what-ai-earbuds-cant-replace-the-value-of-learning-another-language-264965

In this episode we explore the increasingly heated debate over technology in schools — from state-level efforts to restrict devices to the growing pressure on districts to adopt AI tools. Using recent arguments about EdTech, screen bans, handwriting, typing, and classroom technology, we look at why the real question is not simply “screens or no screens,” but what specific learning task students are being asked to do.Topics covered:Why some policymakers are pushing broader school device bansThe debate over whether EdTech harms student learningWhat handwriting vs. typing research actually suggestsWhy learning decisions should be task-specific, not technology-specificHow blanket screen bans can affect students with learning differencesThe difference between banning cell phones and banning instructional toolsWhat teachers should ask when device policies reach the classroomWhat district leaders should consider before adopting or restricting technologyWhy educators, not blanket policies, are best positioned to judge what learning requiresSource:https://www.the74million.org/article/some-states-are-banning-much-more-than-phones-in-schools-thats-a-huge-mistake/https://www.chalkbeat.org/2026/03/17/jared-cooney-horvath-says-ed-tech-hurts-learning-a-look-at-the-evidence/

In this episode we explore why Gen Z students are growing more skeptical of AI even as they continue using it at high rates. Drawing on new Gallup survey findings and recent conversations about “cognitive surrender,” we look at what student attitudes toward AI may be telling educators right now: not that students want AI banned, but that they want real help using it thoughtfully.We also examine what this means for teachers, instructional coaches, and district leaders as schools try to respond to a generation that is both highly engaged with AI and increasingly uneasy about its effects on thinking, originality, and learning.Topics covered:Gen Z’s rising anger and declining excitement about AIWhy AI use is holding steady even as trust fallsThe connection between student skepticism and “cognitive surrender”What students are worried AI may be doing to their thinkingWhy schools should treat student attitudes toward AI as a teaching topicHow teachers can surface skepticism productively in the classroomWhat instructional coaches should look for in AI-integrated lessonsWhy district messaging about AI may need to shiftThe gap between student need and school responseWhat it would look like to help students use AI without hypeSource:https://www.the74million.org/article/gen-z-increasingly-skeptical-of-and-angry-about-artificial-intelligence/

In this episode we explore what happens when AI writing feedback improves a paper without necessarily improving the writer. Using new research from the University of Pennsylvania on graduate students using an AI teaching assistant, we look at the difference between revision and real learning, why group feedback can mask individual understanding, and what teachers can do to make AI feedback more instructionally meaningful.Topics covered:How AI feedback can lead to revision without deep learningWhat a University of Pennsylvania study reveals about student writing improvementWhy group work can hide whether individual students actually learnedThe difference between fixing writing and understanding writingWhy structured reflection should be part of any AI feedback workflowHow teachers can help students process feedback more thoughtfullyWhat district leaders should look for in AI writing-feedback toolsThe risk of overwhelming students with too much unprioritized feedbackWhy revision is evidence of compliance, not always evidence of understandingSource:https://learninganalytics.upenn.edu/ryanbaker/20_U%20Penn%20team%20-%20Do%20Students%20Learn%20from%20Writing%20Feedback%20from%20an%20AI%20Teaching%20Assistant.pdf

In this episode we explore the idea of “cognitive surrender” in classrooms: what happens when students or teachers rely on AI so quickly that they stop fully engaging in their own thinking. Drawing on a new preprint from researchers at Wharton, we examine how access to AI can improve accuracy when the tool is right, but worsen performance when it is wrong, especially under time pressure.Topics covered:Cognitive surrender and what the research suggestsWhy time pressure increases reliance on AIWhich students may be most vulnerable to overtrusting AIWhy AI can make people feel more confident even when it is wrongThe connection between AI use, motivation, and equityWhat teachers can do to design assignments that preserve student thinkingWhy district leaders should connect AI use to workload and staffing conditionsThe role of feedback and meaningful stakes in reducing overreliance on AISources:https://www.nytimes.com/2026/03/29/opinion/ai-claude-chatgpt-gemini-mcluhan.htmlhttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=6097646

In this episode we explore a powerful question for educators: what happens when AI doesn’t just support thinking, but starts to replace it?Drawing on Ezra Klein’s reflections after spending time with people on the frontier of AI, we examine the difference between using AI as a tool for cognitive offloading and slipping into what researchers call cognitive surrender.Topics covered:Ezra Klein’s argument that AI may be changing people, not just technologyThe difference between cognitive offloading and cognitive surrenderWhy polished AI output can conceal shallow or incomplete thinkingWhat this means for student writing, reasoning, and identity formationHow teachers can protect the moments where real thinking needs to happenWhy assignment design matters more in an AI-rich environmentHow instructional coaches can use this conversation with facultyWhat district leaders should consider about the working conditions that make overreliance on AI more likelySources:https://www.nytimes.com/2026/03/29/opinion/ai-claude-chatgpt-gemini-mcluhan.htmlhttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=6097646

In this episode we explore the rise of AI agents in education through the lens of Grammarly’s expanding vision for AI in the workplace and the classroom. We examine the idea that teachers might one day “build” agents that encode their expertise, then ask a harder question: what gets lost when teaching is reduced to a set of rules? We look at what this means for teacher expertise, student equity, and the future of classroom decision-making.Topics covered:What AI agents are, and how they differ from chatbotsGrammarly’s broader vision for agent-based work and learningWhy the idea of “scaling teacher expertise” is both appealing and problematicThe limits of rule-based systems in real classroom instructionWhy pedagogical content knowledge cannot be easily encoded into an agentWhat Teachers Pay Teachers can teach us about quality and scalability in EdTechWhy grammar instruction and history instruction are not the same kind of teaching problemWhy equity concerns should be central when schools evaluate AI agentsThree key questions educators and district leaders should ask before adopting agent-based toolsSources:https://www.theverge.com/podcast/898715/superhuman-grammarly-expert-review-shishir-mehrotra-interview-ai-impersonationhttps://www.gartner.com/en/newsroom/press-releases/2025-08-26-gartner-predicts-40-percent-of-enterprise-apps-will-feature-task-specific-ai-agents-by-2026-up-from-less-than-5-percent-in-2025

In this episode we explore new research on bias in AI writing feedback and what it means for teachers, school leaders, and anyone evaluating AI-powered writing tools in K–12 education. We look at a Stanford preprint on how large language models responded differently to the same student essay when only the attached demographic profile changed, and we connect that to earlier MIT research showing that AI systems can shift tone and quality based on who they think they are talking to. The bigger question: when AI claims to “personalize” feedback, is it actually supporting equity, or quietly automating lower expectations?Topics covered:How a Stanford research team tested AI writing feedback using 600 real eighth-grade persuasive essaysWhy changing only demographic labels changed the feedback students receivedWhat “positive feedback bias” and “feedback withholding bias” can look like in classroom practiceHow AI can give more praise but less useful critique to some studentsWhat the earlier MIT chatbot study revealed about tone, condescension, and perceived vulnerabilityWhy AI “personalization” can slip into profilingWhat teachers should ask before trusting AI-generated writing feedbackHow students can be taught to question, audit, and respond critically to AI feedbackWhat curriculum leaders and district leaders should demand from vendors about inputs, transparency, and equity testingSource:https://arxiv.org/pdf/2603.12471

In this episode we explore what AI can and cannot do when it comes to helping young people prepare for hard conversations. Starting with a story about Gen Z workers using ChatGPT to rehearse salary negotiations and workplace conflict, we contrast that with the “spirit of the game” in ultimate Frisbee, where players are expected to resolve disagreements themselves in real time.From there, we ask a bigger question for schools: if students need AI to simulate difficult conversations, what does that reveal about the kinds of learning experiences we are — and are not — designing for them?Topics covered:How Gen Z is using ChatGPT to rehearse difficult conversationsWhat ultimate Frisbee’s “spirit of the game” can teach us about school designWhy schools may not be giving students enough practice with real disagreementThe difference between AI role-play and real interpersonal fluencyWhat classroom teachers should ask before using AI conversation toolsWhat district leaders should consider when evaluating these toolsWhy civility and conflict resolution have to be designed into learning environmentsThe risks of relying on AI before students have authentic practice in real human situations

We explore the question, what should schools protect and prioritize that AI can never replace?We focus on the idea of the human core, including emotional intelligence, interpersonal connection, ethical judgment, and wellbeing. We also examine what happens when AI takes on more instructional and cognitive tasks, and whether schools are at risk of sidelining the very relationships that matter most for learning.Topics:The Transcend Education framework for holistic student outcomesWhat the report means by the “human core”Why relationships are central to learning, not separate from itHow AI could either strengthen or weaken teacher-student connectionThe risk of personalized learning becoming isolatingWhy schools need to make empathy, ethical judgment, and wellbeing explicit goalsA practical question school leaders should ask when evaluating AI toolsWhy time saved by AI should be redirected toward deeper human workSource:https://transcendeducation.org/wp-content/uploads/2026/01/Overdeck-AI-Resource_FINAL_012026.pdf

We explore one of the more promising classroom uses of AI: helping teachers create custom decodable texts for early readers. We look at how AI can make it much faster to generate phonics-aligned reading passages connected to students’ interests and classroom experiences, while also examining why teacher expertise still matters at every step.Topics:How AI can help generate custom decodable texts in minutesWhy decodable texts matter in early literacy instructionThe difference between generic published passages and classroom-specific reading materialsHow personalization can increase student engagement and relevanceWhy teacher phonics expertise is essential when using AICommon AI mistakes in literacy tasks, including misidentifying sound patternsHow to prompt AI with clear constraints and instructional goalsWhy iteration and revision are part of effective AI useThe “80/20 rule” for using AI drafts and teacher refinementWhy this is a strong low-stakes entry point for educator AI professional learningSource:https://www.ascd.org/el/articles/how-to-create-custom-decodable-texts-in-minutes

We explore what a publishing controversy reveals about AI, authorship, and trust — and why educators should pay close attention. Using the case of a novel reportedly found to be largely AI-generated, we examine how the same challenges facing publishers are also showing up in schools.Topics covered: • The controversy around an AI-generated novel making it deep into the publishing pipeline • Why AI detection remains unreliable in both publishing and education • The difference between detection and attribution • What “fidelity” means in the age of AI-assisted work • Why ambiguity around acceptable AI use creates problems for students and professionals alike • How transparency can reduce misalignment, confusion, and hidden AI use • Why media literacy now has to be an ongoing practice, not a one-time lesson • What teachers and school leaders should take away from this momentSource:https://www.nytimes.com/2026/03/19/books/ai-fiction-shy-girl.html

Drawing on the work of Lee Shulman and a recent interview with researcher Andrew Lan, we look at why AI can generate content and imitate pedagogy, yet still struggle to understand how real students misunderstand, make mistakes, and learn. We also consider what this means for evaluating EdTech tools, teaching AI literacy, and the future potential of simulated student agents in instructional planning.Topics covered: • What pedagogical content knowledge (PCK) is, and why it matters so much in teaching • Why AI still struggles to simulate authentic student thinking and misconceptions • How this helps explain the “jagged frontier” of AI in education • Why many AI-powered education tools may underdeliver in real classroom settings • The difference between knowing content, knowing pedagogy, and knowing how students learn specific content • Why current approaches to AI literacy may focus too much on tools and not enough on underlying understanding • How simulated student agents could eventually help teachers test lessons and prompts before instruction • Why teacher expertise remains difficult to replicate and especially valuable in the age of AISource:https://the-learning-agency.com/the-cutting-ed/article/5-questions-with-andrew-lan/

We explore what kinds of work should teachers actually offload to AI, and what kinds of work should stay fully human?We examine the argument that so-called “busy work” is often not busy work at all, but part of the real craft of teaching. We push on that idea by asking where AI can genuinely reduce friction and where it risks weakening judgment, reflection, and professional voice.Topics:Why some educators argue teachers should not offload “busy work” to AIWhether writing parent emails, designing assessments, and planning instruction are part of the craft of teachingThe risk of letting AI replace teacher reflection instead of supporting itHow AI can reduce friction without replacing the important cognitive workThe difference between tasks where thinking is the point vs. output is the pointWhat schools should consider before encouraging teachers to use AI for efficiencyHow district leaders can distinguish between productive AI use and harmful overrelianceSource:https://www.edweek.org/technology/opinion-why-teachers-shouldnt-offload-their-busywork-to-ai/2026/03

We explore how “AI slop” is flooding children’s media, and why that matters for classrooms, not just screens at home. We look at a new story from The 74 on the rise of AI-generated kids’ videos that appear educational but are often inaccurate, incoherent, or even unsafe.Topics covered:How AI-generated children’s videos can look educational while teaching nonsenseWhy “educational-looking” is no longer the same as genuinely educationalWhat AI slop may do to early learning, comprehension, and concept formationWhy media literacy may need to start earlier and in simpler waysHow teachers should evaluate AI-generated instructional materials for coherence, not just accuracyWhy districts may need clearer policies on AI-generated classroom contentWhat families need to understand about trust, quality, and children’s media in an AI worldSource:https://www.the74million.org/zero2eight/ai-slop-is-flooding-childrens-media-parents-should-be-very-alarmed/https://www.youtube.com/watch?v=f8Z1EmBmf40

We cover a core tension in AI and writing instruction: just because AI can take feedback and grading work off a teacher’s plate, does that mean it should? We explore what’s gained when teachers stay deeply involved in reading and responding to student writing, and where AI may help without replacing the professional judgment that strong writing instruction depends on.Topics covered:Why writing is one of the hardest and most consequential areas for AI in schoolsWhat teachers gain by personally reading and diagnosing student writingThe role of formative assessment in improving writing instructionWhy feedback timing matters so much for student growthThe practical limits teachers face when responding to large volumes of student writingHow AI might help with writing feedback without fully replacing the teacherThe risks of over-automating grading and feedbackBias, inconsistency, and hallucination in AI writing toolsWhat teachers and district leaders should weigh before adopting AI for writingSource:https://www.edsurge.com/news/2026-02-24-what-students-gain-when-teachers-not-ai-grade-students-work

We explore a deceptively simple but increasingly urgent question: How should schools teach students to use AI?We dig into a major report arguing that the risks of generative AI in children’s education may outweigh the benefits. From there, we examine the difference between AI-enriched learning and AI-diminished learning, and we ask what happens when schools put the tools ahead of the thinking.We also look at the risks of a tools-first approach, including cognitive offloading, reduced student agency, and the possibility that schools are teaching students to use AI before helping them understand it. Along the way, we highlight examples from schools taking a more reflective, pedagogically grounded approach.Topics:AI’s risks in education may currently outweigh its benefitsThe difference between AI-enriched learning and AI-diminished learningWhy a tools-first approach can distort how students and teachers understand AICognitive offloading and how AI may weaken student thinking and creativityWhy teaching students to use AI is not the same as teaching them to understand AIWe discuss the importance of prioritizing agency over agility in AI educationWe ask what all of this means for teachers, school leaders, and district decision-makersSource:https://www.washingtonpost.com/opinions/2026/03/10/ai-schools-education-technology-artificial-intelligence/

We explore what it would mean to move beyond using AI to simply speed up traditional tests and instead use it to rethink assessment itself.  We look at how AI could help schools design more authentic, responsive, and instructionally useful assessments.We also explore a central tension: will AI merely help schools build a faster version of the assessment systems they already have, or can it help educators create something fundamentally better for students and teachers?Topics covered:How AI could reshape assessment beyond efficiency and automationWhy assessment should function more like coaching than scorekeepingThe difference between optimizing old systems and innovating new onesWhy authentic performance tasks matter for student learningHow AI might help teachers hear more student thinking, more oftenThe importance of design intent in AI-enabled assessmentWhy routines, infrastructure, and school culture matter as much as the tool itselfWhat “assessment literacy” means for students and teachersThe promise and limits of AI feedback in supporting deeper learningWhy the teacher’s role remains essential, even as AI grows more capable

We take a closer look at one of the most repeated claims in AI and education: that AI is saving teachers hours every week. But instead of accepting the headline at face value, I ask a more important question: what should schools actually be optimizing for?Topics covered in this episode:Why claims about AI “saving teachers time” deserve more scrutinyThe limits of self-reported productivity dataWhy optimizing for efficiency can miss the real point of instructionA classroom example of students using AI feedback on writingWhy AI should function as a thinking partner, not an authorityThe idea of transfer: when AI-supported practice helps students improve independentlyHow structured feedback can build student metacognitionWhy students need to evaluate feedback, not just accept itWhat timely, specific feedback makes possible in writing instructionWhy the best use of AI may be enabling more iteration, reflection, and ownership for studentsWhat educators risk losing when “time saved” becomes the main metricSources:https://www.washingtonpost.com/nation/2026/02/22/ai-chatbots-teach-writing/https://danmeyer.substack.com/p/i-dont-believe-this-finding-that

We explore a provocative question at the center of today’s AI conversation in education: when AI boosts productivity, what actually happens to the time it saves? Drawing on a recent workplace study discussed in Harvard Business Review, we examine whether AI reduces workload or quietly expands it, and what that could mean for teachers, school leaders, and the future of work in schools.Topics covered in this episode:Mark Cuban’s framing of two types of AI usersA workplace study suggesting AI may intensify work rather than reduce itWhy “time savings” do not necessarily mean less workThe three forms of work intensification: task expansion, blurred boundaries, and multitaskingWhat these patterns could mean for teachers and school staffWhether AI efficiency gains actually create meaningful downtimeHow AI tools may begin to co-opt teacher voice and professional judgmentWhy district leaders should think carefully about how work is redesigned around AISources:https://hbr.org/2026/02/ai-doesnt-reduce-work-it-intensifies-ithttps://x.com/mcuban/status/2023750950322889050

We explore a central tension in AI and writing instruction: if educators want authentic student work, is handwriting really the answer?Using research on handwriting versus typing, plus reflections on Chris Lehman’s argument for transparency in schools, we examine why a “return to the blue book” may create new problems instead of solving the old ones.Topics covered in this episode:Why concerns about AI-generated writing are pushing some schools back toward handwritten assessmentsWhat research says about handwriting, typing, memory, and compositionThe difference between writing for learning and writing for composing ideasWhy handwritten work may disadvantage some studentsEquity concerns for students who rely on digital accommodationsThe tradeoffs between preventing AI misuse and preserving authentic writing practiceWhy transparency may be a better long-term strategy than banning toolsWhat Science Leadership Academy’s policy approach can teach schools right nowSources:https://www.scribd.com/document/956588967/Att-I4-HF-Mtnl65BnxCQFqgKkCDxgIJY6v4LHufLYmnqy8https://ies.ed.gov/ncee/wwc/practiceguide/17https://practicaltheory.org/blog/2026/03/07/ai-and-showing-our-work/https://pubmed.ncbi.nlm.nih.gov/15823243/https://psycnet.apa.org/record/2012-04390-002https://psycnet.apa.org/record/2014-35383-001http://frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2023.1219945/fullhttps://www.sciencedirect.com/science/article/abs/pii/S0360131515300920https://www.edweek.org/technology/teachers-turn-to-pen-and-paper-amid-ai-cheating-fears-survey-finds/2023/10

We reflect on Chris Lehman’s recent writing about the state of AI in education and explore a practical middle ground between AI hype and total rejection. We look at why schools are still wrestling with cheating, misaligned expectations, and the real limits of “innovative” lesson design, while also asking where AI may genuinely help, especially in assessment and writing.Topics covered:Chris Lehman’s argument that AI in schools is “not going well”The two camps in AI education: resistance vs. revolutionWhy AI may still hold promise for more dynamic assessmentThe connection between AI-powered feedback and AI-enabled cheatingWhy schools cannot “lesson plan” their way out of AI misuseMisalignment between what teachers assign and what students think the work isWhy transparency with students, staff, and families matters more than everHow quickly students can test, bend, and get around AI guardrailsWhy hands-on experimentation helps school leaders make better AI decisionsSource:https://practicaltheory.org/blog/2026/03/07/ai-and-showing-our-work/

We talk about what happens when district AI guidance arrives after schools have already spent years figuring things out on their own. We look at the newly released New York City Department of Education AI guidance and compare it with how individual schools across the city have already been responding to AI in very different ways.Topics covered:NYC DOE’s newly released AI guidanceHow schools responded before district policy existedThe difference between proactive, reactive, and passive school responsesWhy delayed policy creates inequity across classrooms and campusesThe challenge of defining acceptable vs. unacceptable AI useWhat students are telling us about school’s role in preparing them for an AI-enabled worldWhy assignment clarity matters more than everThe tradeoffs between paper-and-pencil writing and digital writingWhat school leaders can learn from values-based AI policy designWhy professional development alone may not be enoughSources:https://www.schools.nyc.gov/about-us/vision-and-mission/guidance-on-artificial-intelligencehttps://www.chalkbeat.org/newyork/2026/03/23/schools-develop-ai-policies-awaiting-city-guidance/

In this episode we unpack New York City Public Schools’ new red, yellow, green AI guidance framework and explore what it signals for teachers, school leaders, and district administrators across the country.We look at why this policy matters not just because NYC is the largest school district in the U.S., but because it offers one of the clearest examples yet of how a district is trying to separate prohibited AI uses, human-in-the-loop uses, and approved professional uses. We also discuss what schools can learn from the parts of the policy that draw hard boundaries around grading, discipline, placement, graduation decisions, IEPs, and student data privacy.Topics covered:NYC’s new red, yellow, green framework for AI use in schoolsWhy AI cannot be used for placement, discipline, promotion, or graduation decisionsWhy IEPs and 504 plans must remain in the hands of qualified professionalsThe importance of protecting student data from AI model training and monetizationWhat it means to keep human judgment in the loopWhy student AI use falls into a yellow-light categoryWhere teachers have a green light to use AIWhat instructional coaches can do to help teachersWhat district leaders should review immediately for liability and policy complianceWhy professional learning will matter if schools want staff to understand the difference between green, yellow, and red-light usesSource: NYC DOE

Who is classroom tech really built for?We reflect on a long-running tension in K–12 EdTech: the gap between what districts buy, what teachers are expected to use, and what students actually experience. We explore what happens when software is designed more for dashboards, monitoring, and accountability than for genuine teaching and learning.Topics covered in this episode:Why critiques of EdTech in 2026 sound a lot like critiques from 2010The difference between building for the buyer versus building for the classroom userHow district purchasing incentives can shape product designWhy student engagement is often missing from EdTech adoption decisionsWhat it means for a teacher to have a clear role during technology-supported instructionA contrasting example of when classroom software actually deepened student learningHow product philosophy and go-to-market strategy can either align or conflictWhy educators should ask whether a tool serves learning, accountability, or both

We examine the hidden costs of “free” AI tools in schools. The real issue is not just convenience or productivity. It’s what happens when teachers upload student writing, IEP notes, and other sensitive information into tools that districts have not approved and may not fully understand.Topics covered in this episode: • Why “free” AI tools are often paid for with user data • How student writing can become unrecoverable once uploaded to AI systems • The risks of using AI to generate IEPs and student support plans • Why shadow AI is becoming a serious district-level threat • The tension between teacher efficiency and slow-moving district policy • How unapproved AI use can expose districts to lawsuits and liability • Why teachers and administrators need stronger AI literacy • What districts can do to “ring-fence” approved AI tools and coach staff more activelySource:https://thejournal.com/articles/2026/02/25/shadow-ai-is-quietly-becoming-k12s-biggest-cybersecurity-risk.aspx

In this episode, we examine a troubling MIT study showing that AI chatbots can respond differently based on who they think they’re talking to, sometimes giving less accurate, more patronizing, or even mocking responses to users described as less educated or non-native English speakers.This episode looks at:what AI bias can look like in practice why district leaders need better ways to evaluate these toolshow researchers and schools might work together on accountability why AI alignment is not a finished problem what it means when AI systems “get to know” users over time

We take a closer look at Humble ISD’s AI pilot with OpenAI and ask the deeper questions that often get missed in early conversations about district AI adoption.Topics covered in this episode: • Humble ISD’s pilot of ChatGPT for teachers • Why “time savings” is not enough as a success metric • What districts should measure beyond efficiency • Whether AI-created time actually leads to better student support • The difference between implementation success and instructional success • Risks of teacher overreliance on AI tools • How AI could affect teacher agency and lesson planning • Why districts need guardrails, transparency, and community discussion • The importance of aligning AI pilots to district mission and values • What to do if AI pilots show only marginal gains in student outcomes

Nearly 6 in 10 teens say AI cheating is happening at their school. But what if the real issue is not just cheating, it’s confusion about what schoolwork is actually for?We explore a powerful idea from Tony Frontier’s AI with Intention: fidelity. When students use AI to summarize a text, generate ideas, or speed up part of an assignment, are they cheating, or are they responding to unclear expectations about the purpose of the work?Topics coveredPew research on AI cheating in schoolsFidelity vs. cheatingAssignment purpose and teacher communicationAI as a shortcut vs. AI as a supportReading comprehension, writing, and hidden learning lossWhy explicit expectations matterImplications for classroom practice and district policyPew researchAI with Intention: Principles and Action Steps for Teachers and School Leaders

We explore Google’s new AI-suggested feedback feature in Google Classroom and ask the bigger question: will tools like this actually improve writing instruction, or just make it easier to automate teacher comments?Topics covered:Google’s new AI writing feedback in Google ClassroomWhether AI can reduce teacher workload in meaningful waysWhy rubric-based feedback may matter more than chatbot-style interactionThe risk of teachers “rubber stamping” AI suggestionsHow AI could influence the authenticity of teacher-student communicationSource:https://workspaceupdates.googleblog.com/2026/02/educators-now-get-help-drafting-personalized-guidance-on-written-assignments-with-AI.html