← Back to Blog
🎓 AI in Education

AI in Education Digest — November 9, 2026

Dark navy graphic with the white serif headline Who does, and below it in teal, the thinking?, underlined in teal. On the right, a lined notebook with three handwritten-style white lines and a teal pencil resting on the page. Footer: AI in Education Digest, 9 Nov 2026, derecktafuma.co.uk.

Seven stories, one question: who does the thinking? In England, 300+ experts want a pause on AI tutors for disadvantaged pupils. In the US, teachers are moving work back into the classroom with oral exams, the University of Chicago is making an AI-free writing course compulsory, and Seattle closed a back door to Gemini hidden inside Google Search. Liverpool City Region is training every school leader, Australian universities are labelling each task Can, Can’t or Must use AI, and the Philippines is teaching children to set boundaries with chatbots. The takeaway for families: decide task by task, and keep the first attempt human.

1. 300+ teachers, parents and doctors ask ministers to pause England’s AI tutor trial

Source: Sophie Huskisson, The Mirror · 8 November 2026

More than 300 teachers, parents, academics and doctors have written to Andy Burnham and Education Secretary Lucy Powell asking them to halt plans to give AI tutoring tools to up to 450,000 disadvantaged children from next year. The letter says these children “should not be the first to bear the risks of an unproven and contested technology”, and that hunger, absence and mental health “are not technological problems”. Signatories include the SafeScreens Campaign, Health Professionals for Safer Screens and Schools Beyond Screens. The government says it is “not a national rollout of AI tutors, but a carefully controlled trial” to test safety before anything is scaled up.

Why it matters: This is the big UK question for families: is an AI tutor a fair substitute for a human one? The trial’s evidence will shape what schools are offered, so it’s worth watching what gets measured.

X reaction: @TheShallowEndAI (8 Nov, replying to the reporter): “If schools hit pause on AI tutors, families can still keep a simple rule at home: use it like a study partner that asks questions, not a service that writes the assignment.”

Video angle: “450,000 children could get an AI tutor next year. 300 experts say: not yet.”

2. Teachers swap take-home essays for oral exams and in-class writing

Source: Helen Coster, Reuters (via Arizona Daily Star) · 7 November 2026

Reuters interviewed a dozen US teachers and experts and found a clear shift: less take-home work, more in-class writing, discussion and face-to-face assessment. A San Francisco sixth-grade teacher dropped take-home essays for five-minute oral exams of 15 questions; a New York maths teacher now sets fewer homework questions and more in-class practice so he can see how students work; another English teacher spends a month writing one essay with his class.

Why it matters: If homework can be done by a chatbot, the test moves to the room. For GCSE students that means the skills that count are the ones you can show without help: explaining, recalling and writing under time pressure.

X reaction: @think1more (7 Nov) summarised three university responses: “Yale: in-person exams up from 14.8% to 20% … Hebrew U: oral exams, paper presentations”.

Video angle: “Teachers are bringing back the oral exam. Here’s why AI made it happen.”

3. University of Chicago will make an AI-free writing course compulsory for first-years

Source: Kathryn Palmer, Inside Higher Ed · 9 November 2026

UChicago has launched the Chicago Center for Writing, and from next autumn every first-year student must take an AI-free writing course run through it. The same roundup reports students at Eastern Illinois University petitioning for rules on how much professors rely on AI for course materials (529 signatures), and a Berkeley tool, Rubrica, that grades handwritten exams against a rubric, matching human scores 85% of the time in a small study, with the instructor reviewing every grade.

Why it matters: Top universities are signalling that writing on your own still matters. Students heading for university should practise drafting without AI now.

X reaction: No X reaction found for the UChicago course in this window (searched 7–9 Nov).

Video angle: “One of America’s top universities just banned AI from a compulsory class.”

4. Seattle schools close a back door to Gemini through Google Search

Source: The Seattle Times · 7 November 2026; background: Gothamist · 6 November 2026

Seattle Public Schools already blocks chatbots such as Claude and Grok, but students found that clicking the AI summary in Google Search opened a Gemini chat. From this week a web-filter update closes that route, alongside blocking YouTube for K–5 and Minecraft (except where teachers use it), and a 60-minute daily screen cap for grades 3–5. The district hasn’t banned generative AI: approved tools such as MagicSchool AI, Canva Magic Write and Adobe Firefly stay. In New York, Gothamist reports the city has disabled Gemini and AI Mode on student Chromebooks and asked Google to remove AI Overviews, after last week’s row over Google tying Search to AI.

Why it matters: It follows last week’s Google story: schools are learning that “AI off” needs checking, because AI now sits inside ordinary tools like Search.

X reaction: @SEATTLESUBMISS (9 Nov, about 1.4k views) shared the change as “Seattle Public Schools to Block Minecraft and Google Gemini for All Students, Along With YouTube for Grades K–5”.

Video angle: “Students found a secret door to an AI chatbot inside Google Search. Seattle just shut it.”

5. Liverpool City Region opens free AI training to every school leader

Source: Labour Home · 8 November 2026; Mersey Monitor (week of 2 Nov)

The Liverpool City Region Combined Authority announced its largest regional rollout of the Big AI Project: free training, resources and guidance on safe, ethical AI use for headteachers, trust leaders and staff, designed by teachers through the Oasis Big Education Institute and described as technology-neutral. Earlier the same week it expanded a Century Tech personalised-learning platform to 20 secondary schools and more than 3,000 pupils, after a primary pilot with nearly 5,000 children. Mersey Monitor notes Century Tech’s chair is Lord Jim Knight, a Labour peer and former schools minister.

Why it matters: Local areas are moving faster than national policy. Parents in the region can ask their school whether it has joined, and which tools it uses.

X reaction: No X reaction found in this window (not searched separately).

Video angle: “A whole UK region just offered every head teacher free AI training. Here’s what’s in it.”

6. Australia: four universities widen ChatGPT Edu, with “Can, Can’t, Must” rules for assessments

Source: Mara Sugue, ITBrief · 9 November 2026

OpenAI has expanded partnerships with UNSW Sydney, the University of Sydney, La Trobe and Adelaide University. UNSW is extending ChatGPT Edu to more than 80,000 students and staff, and has a Can, Can’t, Must framework where each assessed task says whether AI is permitted, prohibited or required. Sydney assesses unaided knowledge in supervised settings and AI use in open tasks. An OpenAI-commissioned survey of 1,018 Australian students found 81% use AI several times a week and 33% several times a day.

Why it matters: This is the clearest model yet for fair AI rules: decide per task, not with a blanket ban. Families can use the same idea for homework at home.

X reaction: @eduitguy (7 Nov): “To me it’s unis moving to managed access with clear rules, a shift worth school IT watching.”

Video angle: “Can, Can’t, Must: the simplest AI rule for homework I’ve seen.”

Navy board titled AI rules, task by task. Three cards side by side: Can, with a teal switch on; Can't, with a switch off; Must, outlined in teal with a teal switch on. Footnote: UNSW's assessment framework, teachers mark each task as AI permitted, prohibited or required. Source: ITBrief, 9 Nov 2026.

UNSW’s Can, Can’t, Must assessment framework. Source: ITBrief, 9 Nov 2026.

Navy bar chart titled Uni students using AI. Two teal bars: Weekly 81 percent, Daily 33 percent. Footnote: weekly means several times a week, daily means multiple times a day, 1,018 Australian university students. Source: OpenAI-commissioned survey reported by ITBrief, 9 Nov 2026.

How often Australian university students use AI. Source: OpenAI-commissioned survey via ITBrief, 9 Nov 2026.

7. Philippines: Microsoft and DepEd to train 2,000 teachers in AI and chatbot safety

Source: Jinji Quiambao, NewsBytes.PH · 7 November 2026

Microsoft and the Department of Education will train and credential 2,000 public school teachers in AI literacy, with sessions from November to May. Alongside Microsoft Elevate for Educators and an AI Classroom Toolkit, a student-safety resource called Behind the Chat uses fictional teen scenarios to cover privacy, oversharing, AI-assisted decisions and emotional attachment to chatbots, with guidance for parents too.

Why it matters: Emotional attachment to chatbots is now part of safety teaching, not just cheating. That’s a conversation worth having at home whatever country you’re in.

X reaction: No X reaction found in this window (not searched separately).

Video angle: “A country is teaching kids not to get attached to chatbots. Should we?”

AI in educationAI tutorsassessmentoral examsacademic integrityGoogle Searchschool leadersAI rulesonline safetyrevision

Want this digest in your inbox? A short roundup of what’s actually changing in AI and learning.

Join the newsletter →

Keep the first attempt human

Get The Revision Lab’s free GCSE History revision sheet: key people and dates for Black History Month on two A4 pages, with a timeline, exam tips and a quick quiz with answers, so your child can test themselves with no AI. The GCSE Maths Mock Booster is live.

Get the free revision sheet The Revision Lab