[{"data":1,"prerenderedAt":87},["ShallowReactive",2],{"blog-post-\u002Fblog\u002Fai-marking-lesson-planning-real-vs-hype":3,"blog-surround-\u002Fblog\u002Fai-marking-lesson-planning-real-vs-hype":80},{"id":4,"title":5,"author":6,"body":7,"date":64,"description":65,"draft":66,"extension":67,"image":68,"imageAlt":69,"meta":70,"navigation":71,"path":72,"seo":73,"sitemap":74,"stem":75,"tags":76,"__hash__":79},"blog\u002Fblog\u002Fai-marking-lesson-planning-real-vs-hype.md","AI marking and AI lesson planning: what's real, what's hype, and what teachers are actually trying","Classroom Hub team",{"type":8,"value":9,"toc":57},"minimark",[10,14,26,31,34,37,41,44,47,51,54],[11,12,13],"p",{},"Two things are true at once about AI in schools right now, and most coverage only manages to capture one of them. The first is that adoption among teachers has risen genuinely fast — from roughly half to more than three-quarters in the space of about a year, which is a startling rate of change for any tool in any profession. The second is that almost none of that growth is happening in the place the public conversation assumes it is. Marking — the use case that dominates the headlines and the anxieties — is, by a clear margin, the area where teachers trust these tools least.",[11,15,16,17,21,22,25],{},"That gap between ",[18,19,20],"em",{},"where the conversation is"," and ",[18,23,24],{},"where the actual use is"," is worth pulling apart properly.",[27,28,30],"h2",{"id":29},"where-the-real-growth-is-happening","Where the real growth is happening",[11,32,33],{},"The vast majority of day-to-day AI use by teachers sits in distinctly unglamorous territory: drafting resources, generating starter activities, producing differentiated versions of a worksheet, chipping away at the kind of admin that used to eat an evening. None of that makes for a dramatic headline. All of it is the sort of small, repeated task that — multiplied across a week — adds up to real time back.",[11,35,36],{},"This is, in a way, the least surprising place for adoption to take off first. These are low-stakes uses: if a generated starter is mediocre, a teacher edits it or binds it in five minutes, and nothing downstream depends on it being perfect. The cost of a bad output is small, and the time saved on a good one is real. That's exactly the kind of use case that spreads fast through word of mouth in a staffroom — not because anyone announced a strategy, but because it quietly works often enough to be worth trying again.",[27,38,40],{"id":39},"where-the-trust-isnt-there-and-why-thats-not-irrational","Where the trust isn't there — and why that's not irrational",[11,42,43],{},"Marking sits at the opposite end of that spectrum, and teachers' caution about it isn't technophobia — it's an accurate read of where the stakes actually are. A mediocre AI-generated worksheet costs five minutes to fix. A wrong or inconsistent piece of AI-generated feedback on a student's work can shape how that student sees their own ability, can misinform a parent conversation, and can sit in a record that follows the student for months. The asymmetry between \"low cost of being wrong\" and \"high cost of being wrong\" maps almost exactly onto where adoption has and hasn't taken hold — which suggests teachers, as a profession, are making a fairly sound collective judgement about where the tool is ready and where it isn't.",[11,45,46],{},"It's also worth noting that a meaningful share of teachers using these tools say they're doing so without any clear school policy on how. That's not really a story about teachers moving recklessly — it's a story about guidance lagging a year or more behind practice, which is a familiar shape for any fast-moving technology to take inside a slower-moving institution.",[27,48,50],{"id":49},"the-honest-middle-ground","The honest middle ground",[11,52,53],{},"None of this adds up to either of the two extreme stories that tend to circulate — \"AI is going to replace the marking\" or \"AI has no place near student work.\" The real picture is narrower and more interesting than either: a genuinely useful tool for the unglamorous, low-stakes, repeatable parts of the job, used carefully and with active scepticism in the parts where being wrong actually costs something.",[11,55,56],{},"That's not a dramatic conclusion. It's also probably the most accurate one available right now — and a useful thing to keep in mind the next time a headline tries to tell you AI has either solved or ruined some part of the profession overnight. The detail, as usual, sits somewhere quieter in the middle.",{"title":58,"searchDepth":59,"depth":59,"links":60},"",2,[61,62,63],{"id":29,"depth":59,"text":30},{"id":39,"depth":59,"text":40},{"id":49,"depth":59,"text":50},"2026-05-11","Adoption has jumped sharply in the past year — but not in the places the headlines focus on. A look at where AI is actually earning a place in teachers' workflows, and where it isn't, yet.",false,"md","\u002Fblog\u002Fog\u002Fai-marking-lesson-planning-real-vs-hype.svg","AI marking and AI lesson planning: what's real, what's hype, and what teachers are actually trying — Classroom Hub",{},true,"\u002Fblog\u002Fai-marking-lesson-planning-real-vs-hype",{"title":5,"description":65},"[object Object]","blog\u002Fai-marking-lesson-planning-real-vs-hype",[77,78],"education-news","edtech","JxNOjoMqHJj6js2HL6QfhxTtOqqrKDV7Tlii_QEpjtw",[81,82],null,{"title":83,"path":84,"stem":85,"description":86,"children":-1},"Attendance registers: the two-minute task that becomes a Friday afternoon scramble","\u002Fblog\u002Fattendance-registers-friday-afternoon-scramble","blog\u002Fattendance-registers-friday-afternoon-scramble","Taking the register only takes a moment. Making sense of a week of registers, alongside everything else you tick off at the door each morning, is a different job entirely.",1780939452607]