This page is a real example of AIMingUp!, not a brochure. A genuine student lesson exactly as a child would receive it, the standard topic list a school or a group can select from, and the research the whole thing is built on, with every source named so you can check it yourself.
Both are unedited. Both are delivered by an AI digital twin, labelled as one on screen throughout, because we are teaching AI honesty and a hidden trick would be the wrong note to strike with a school. The two are deliberately different, because a fifteen year old and an eighteen year old are not facing the same risk.
Lower secondary. AI can now fake a person who exists, and invent one who does not. Both are named in the September 2026 safeguarding update as matters in their own right.
One spine, two halves. Faking a person who exists, meaning images and voices of a real classmate or teacher. And faking one who does not, meaning a chatbot that behaves like a friend. Recognise, refuse, report.
AI can copy a person. Being one is a different thing entirely. If it involves a real person, or if it feels like a person, stop and check with a real human being you trust.
Every lesson has a one page student action sheet and a teacher note. Nothing is handed out as a loose PDF; students are enrolled so the school can see who has completed what.
Sixth form. Repeating the lower secondary lesson here would insult them. At this age the real risk is quietly outsourcing the thinking they are about to be examined and interviewed on.
Every task produces two things: the work you hand in, and what changes in your head. AI can produce the first. It can never produce the second. An essay you did not write is an essay you cannot discuss.
AI can hand in your work. It cannot sit your interview. Students get a sixty second test they can run on themselves, and three questions to ask before they start any task.
University tutors, interviewers and employers all test what a student actually knows. The lesson also lands the one rule that never fails in any exam board: declare what you used.
Four student lessons split by age so nobody is talked down to, plus the first teacher briefing. Each runs under three minutes. The student lessons stop partway so the class does the exercise before the video goes on. Every lesson opens by saying it is Darryl's digital twin standing in for him. Use them, copy them, run them in your own room. No email address asked for.
Student lesson · Years 8 to 10 · 2 min 22
Student lesson · Years 11 to 13 · 2 min 54
Teacher track · Staff briefing · 2 min 29
Student lesson · Years 8 to 10 · 2 min 49
Student lesson · Years 11 to 13 · 2 min 33
Wave Two, lessons six to twelve, follows the same way. The founders list, the money and the rules are public on the open book page.
A school group does not want a generic AI course with someone else's logo on it. AIMingUp! is built to your programme, your values and your focus, and the topic list on the next tab is where that conversation starts, not where it ends.
Which topics run, in which year groups, in what order and at what depth. Which of your subjects the AI work sits inside. Your safeguarding and acceptable use wording. Your academic integrity rules. Your assessment policy. Your values framing, including a faith based framing where that is who you are.
Your branding on every lesson, worksheet and certificate. Your own domain and page. Your wording on the downloadable sheets. Lessons re recorded where your context needs different examples. A version for each school in the group, or one shared version, whichever suits your structure. And, if you want it, lessons delivered by one of your own staff rather than by Darryl.
Decide once, deploy across the network. One policy, one standard, one set of materials, with room for each school to adapt locally. The strongest argument for a group level decision is consistency: a family moving between two of your schools should meet the same standard of AI teaching in both.
For Catholic and Christian groups, the ethics module can be built on your own tradition's language rather than generic technology ethics. That work draws on published texts, not our opinions, and we name them on the Research tab so your leadership can check the grounding.
The sample lesson is delivered by a digital twin of Darryl. That is the quickest route, but it is not the only one, and for some schools it will not be the right one.
Building a twin of one of your own people is a one off piece of setup, and we would be straight with you about what it involves before you commit to it. Once it exists, it costs no more to run than ours does. Consent, and what happens if that person later leaves, are written into the agreement rather than left to good intentions.
This is the menu schools select from. It is drawn from the frameworks that education systems are actually using, not from our own preferences: UNESCO's two AI competency frameworks, the joint European Commission and OECD AI Literacy framework, the Australian Framework for Generative AI in Schools, England's Department for Education guidance and product safety standards, and the assessment rules set by JCQ and the IB. The tag on each topic shows where it comes from.
Every figure below is attributed to a named study with the sample size, the date and a link. Where the evidence is weak or missing, we say so rather than filling the gap. If any of this changes, we would rather you found out from us than from a parent.
Nine per cent of school leaders reported having an agreed AI strategy, with a further 31% saying one was in development. 75% said there was insufficient AI expertise within their organisation. Their ranked concerns were plagiarism and malpractice 65%, adequacy of training 62%, quality control 58%, data privacy and security 57%, bias and fairness 54%, and safeguarding 44%.
Browne Jacobson School Leaders Survey, February 2025. Over 200 leaders representing roughly 1,650 schools and one million pupils in England. SourceSeven widely used AI detectors were tested on 91 TOEFL essays written by Chinese students and 88 essays by United States eighth graders. The average false positive rate on the second language writers' essays was 61.22%. On the native speakers' essays it was 5.19%. All seven detectors wrongly flagged 18 of the 91 TOEFL essays, and at least one detector flagged 97.8% of them. The same study found that asking a chatbot to rewrite text in more literary language dropped detection from 100% to 13%.
Liang, Yuksekgonul, Mao, Wu and Zou, Patterns (Cell Press), 2023. Full paper. For an international school group, where most students write in a second language, this is the single most important finding on this page.Fourteen detection tools were tested across 54 documents in six categories, 756 tests in total. The best scored 81%, the worst 33%, and the average was around 65%. On text that had been machine paraphrased, accuracy fell to 26%. The authors concluded that the systems tested "should not be used in academic settings". Turnitin's own published position is that its score "should not be used as the sole basis for action", that it does not make a determination of misconduct, and that the technology "might be biased against particular subject areas or second language writers".
Weber-Wulff and colleagues, International Journal for Educational Integrity, 2023. Full paper. Turnitin's own statement: Turnitin FAQs.18% of students reported being accused of unauthorised AI use without proof, or being accused and later exonerated. 24% said they or someone they know lost credit. Teachers reported these events at roughly half the rate students did, which means the problem is systematically under counted at leadership level. 43% of teachers said they regularly use AI detection tools. Separately, only 21% of teachers said their school had shared any policy addressing AI generated intimate imagery, and student receipt of guidance on it fell from 46% to 40% year on year.
Center for Democracy and Technology, "Hand in Hand", October 2025. 806 teachers in grades 6 to 12, 1,030 students in grades 9 to 12 and 1,018 parents, fielded June to August 2025. Source86% of children aged 9 to 17 use or interact with AI, and 85% of those users use it for homework help. But only 35% correctly understood that AI cannot tell whether something is true, and 42% wrongly believed AI can verify accuracy. 17% of chatbot users had encountered content that was not appropriate for their age, and of those, 53% told no adult. 41% had never had a conversation with a parent about AI safety.
Common Sense Media, "AI Use by Tweens and Teens 2026". 1,204 children aged 9 to 17, fielded 18 to 26 March 2026, margin of error 3.8 points. Full reportMore than 80% of students said their teachers had never explicitly taught them how to use AI for schoolwork, while 54% of students and 53% of English, maths and science teachers reported using AI for school, both up more than fifteen points in a year. 45% of principals reported having any school or district AI policy or guidance. 61% of parents worried that AI undermines critical thinking, against only 22% of district leaders, a large gap between what families fear and what leadership believes.
RAND, with the Center on Reinventing Public Education and the MIT Teaching Systems Lab. Nationally representative United States panels of teachers, school leaders, district leaders, students and parents, fielded spring 2025, published 30 September 2025. Source82% of teachers had received no formal guidance on applying AI tools to their work. Separately, the proportion of teachers receiving no AI training at all fell from about 60% in October 2024 to 42% by winter 2026, but only 9% receive ongoing training rather than a one off session. In England, around 20% of schools have a formal AI policy, rising to 34% in secondaries, and staff confidence with education technology generally has fallen, not risen.
Gallup and the Walton Family Foundation, 2,069 United States teachers, February to March 2026. Reported here. Training trend: EdWeek Research Center, 651 teachers, February to March 2026. Source. England policy and confidence data: Department for Education Technology in Schools Survey 2024 to 2025, 1,634 schools, published November 2025. ReportIn the only randomised controlled trial we could find, teachers given a chatbot together with a structured guide cut Key Stage 3 science lesson planning from 81.5 minutes to 56.2 minutes a week, a 31% reduction. An independent teacher panel found no noticeable difference in resource quality, though the trial's authors cautioned that this rested on a limited sample and should be treated with care. Separately, teachers using AI weekly reported saving 5.9 hours a week, about six weeks across a school year.
Education Endowment Foundation with NFER, published 12 December 2024. 259 teachers across 68 English secondary schools. Source. Hours saved: Gallup and the Walton Family Foundation, 2,232 United States teachers, March to April 2025. Source. This finding is why our teacher track is a structured programme and not a tour of tools.Keeping Children Safe in Education 2026 comes into force on 1 September 2026 for schools in England, which at the time of writing is four weeks away. It redefines "nudes and semi nudes" to include images digitally altered or wholly generated using AI, brings generative AI applications that simulate human interaction inside the contact risk category, and requires that filtering and monitoring effectiveness be reviewed at least annually, led by a named senior leader with the designated safeguarding lead, with formal records of what was checked and what was remedied. For any group operating English schools this turns AI from a teaching question into a compliance duty with a date on it.
Keeping Children Safe in Education 2026, Department for Education. Cross referenced to the DfE's Generative AI product safety expectations, updated 19 January 2026. Safety expectationsResearchers at Stanford Graduate School of Education, who have surveyed students anonymously for around fifteen years, report that 60 to 70% of students admit to at least one cheating behaviour in the previous month, and that this figure stayed about the same or decreased slightly after chatbots became widely available. The honest reading is that AI changed the method and the visibility of academic dishonesty far more than it changed the amount. We include this because it argues against panic, and a programme sold on panic is a programme that will not last.
Denise Pope and Victor Lee, Stanford Graduate School of Education. Source. Sample sizes are not published in the public write up, so treat the figure as directional.
UNESCO, AI competency framework for students, 2024. Four aspects, human centred mindset, ethics of AI, AI techniques and applications, and AI system design, across three levels, Understand, Apply and Create, giving twelve competency blocks.
UNESCO, AI competency framework for teachers, 2024. Five aspects across three levels, Acquire, Deepen and Create, giving fifteen competencies.
Empowering Learners for the Age of AI, the AILit Framework, European Commission with the OECD, 2026, approved by the PISA Governing Board on 8 April 2026. Four domains, Engage, Create, Manage and Shape, with nineteen competencies. The most detailed school level source available.
Australian Framework for Generative AI in Schools, approved by Education Ministers on 5 October 2023. Six principles and twenty five guiding statements, the clearest governance checklist of the group.
Generative artificial intelligence in education, England's Department for Education, last updated 12 August 2025, with support materials refreshed 19 May 2026.
JCQ, AI Use in Assessments, April 2025. Binding on any school entering candidates for GCSE, IGCSE or A level, including international centres.
International Baccalaureate, AI in learning, teaching and assessment. Students must credit AI generated work in both the body text and the bibliography.
Singapore Ministry of Education, AI in Education, within the EdTech Masterplan 2030. Learn about AI, learn to use AI, learn with AI, learn beyond AI.
Council of Europe, draft Recommendation on AI Literacy, approved by the Steering Committee for Education in March 2026. Human, technological and practical dimensions. Formal adoption not confirmed at the time of writing, so we cite it as approved rather than adopted.
Antiqua et Nova, a Note on the relationship between artificial intelligence and human intelligence, Dicasteries for the Doctrine of the Faith and for Culture and Education, released 28 January 2025. Its central line for schools is that the physical presence of a teacher creates a relational dynamic that AI cannot replicate.
Magnifica Humanitas, encyclical of Pope Leo XIV, 25 May 2026, which calls for an educational alliance and for media literacy among young people.
Leading Insights: Artificial Intelligence, Association of Christian Schools International, 2024. Includes the distinction between what technology makes possible and the direction human choice takes it, and the concept of latent persuasion, AI quietly shaping a worldview without the user noticing.
Higher Powered Learning, Alliance for Catholic Education, University of Notre Dame. Free policy templates written for diocesan and school system leaders.
The whole course is thirty lessons for US$381. Or start with the first ten for US$147. One child. Paid once. Nothing renews.
A briefing shows your leadership how the programme runs, how it stays safe, what a pilot in two or three schools looks like, and what it would cost. No obligation, and we will tell you plainly if we are not the right fit.
What happens to what you type here: your name, school and email come to Darryl Reeves at CatalystPro.Ai and are used only to reply to you about AIMingUp! They are not sold, not shared and not added to any mailing list. You can ask us what we hold, correct it, or have it deleted, at info@aimingup.org. Full detail on our privacy page.
