Hey {{ first_name | human }},
This week, we’re asking where to draw the line with AI in schools. We look at Ofsted’s expectations, what PISA can—and cannot—tell us about AI and learning, and New York’s restrictions on pupil use. The question running through it all: where can AI help, and where do pupils need to do the thinking themselves?
TL;DR: The 60 Second briefing
⚡️Ofsted & AI: Its position is pretty clear on the use of AI during inspection and it is all down to leaders being able to articulate their use!
🧪What can PISA tell us about AI?: Across OECD countries, 46% of 15-year-olds said they used AI chatbots at least weekly “to help me learn”. But students who used AI for specific school tasks scored lower in science than those who did not.
🚨NY restricts AI: The US school district have moved to restrict most pupil use of generative AI, particularly for younger pupils. New York’s policy limits generative AI for pupils through eighth grade.
📚 AI+education news
⚡️ Ofsted & AI > What it is: I am not sure how new this is, but I think it worth mentioning. In the School inspection operating guide for inspectors, this is what they had to say on the use of AI.
“You do not need to ask if artificial intelligence (AI) is used or actively search for it. If you come across instances where a school uses AI, or has an explicit policy on pupils using AI, record how leaders have chosen to implement it, its impact (for example, on tackling absence), how it is monitored, and the school’s checks and balances to make sure it is accurate, safe and used in the best interests of pupils. If you have any concerns about the school’s use of AI, contact the duty desk.”
Why this matters: The position seems pretty clear. You do not need a separate AI strategy for Ofsted, but you do need to make defensible and smart decisions around its use.
🧪What can PISA tell us about AI? > What it is: PISA 2025 gives us one of the clearest international snapshots yet of how 15-year-olds are using AI for schoolwork.
Across OECD countries, 46% of pupils use AI chatbots at least weekly to help them learn. But pupils who use AI for specific tasks such as summarising texts, drafting writing or carrying out preliminary research generally score lower in science than pupils who do not use AI for those tasks. The average gap is around 20 PISA points.
The picture is more nuanced for general learning. Weekly users perform at a similar level to non-users once socio-economic background is taken into account. Pupils who regularly use AI and are also taught how to assess AI-generated information tend to perform slightly better than frequent users who receive no such guidance.
Why it matters: PISA is observational, so it cannot show that AI caused lower attainment, but the pattern should make us cautious about using AI to replace the thinking involved in reading, researching and writing.
🚨NY restricts AI > What it is: New York City Public Schools has introduced a much more restrictive approach to generative AI for 2026–27.
Pupil-facing generative AI is now banned from pre-K through Grade 8. In Grades 9–12, pupils can use only approved and vetted AI programmes for limited purposes. Exceptions remain for assistive technologies and other necessary supports.
The district says the policy is intended to ensure that technology serves learning rather than replacing the thinking pupils need to do themselves.
Why it matters: New York joins a growing number of education systems questioning the assumption that younger pupils need direct access to generative AI. This feels less like an isolated policy choice and more like an emerging divide in AI policy: teacher use may expand while pupil-facing use becomes more tightly controlled. Schools may well need to decide which side of the divide they align themselves to.
🌍 Wider AI updates
🧪 ChatGPT Astra > What it is: AI gets more agentic.
OpenAI has released GPT-6 Astra, its new frontier model, with stronger performance in areas such as computer use, browsing, coding, science and professional work.
The more significant shift is not just that the model is “smarter”. It is increasingly designed to act across tools and complete multi-step work, rather than simply respond to a prompt. OpenAI is also pushing this through ChatGPT Work.
Why it matters: The next phase of AI is less about generating an answer and more about carrying out a task. For schools, that could eventually mean AI systems that search files, analyse data, create resources and complete administrative workflows with much less human prompting.
‘Till next week.
Mr A 🦾
Help a colleague save time by sharing this newsletter; distributing these ideas helps a friend get home on time and keeps our energy focused on what matters most: great teaching.
Safety & Privacy Notice
The tools and workflows mentioned are intended for professional productivity and educational enhancement. Users must ensure that any AI implementation remains compliant with their local data protection regulations and institutional safeguarding policies.
Data Privacy: Do not enter personally identifiable information (PII), sensitive student records, or confidential institutional data into public AI models.
Verification Required: AI-generated content can be inaccurate, biased, or out of date. Always maintain a "human-in-the-loop" approach by reviewing and fact-checking all outputs before use.
Professional Judgement: These suggestions do not substitute for formal legal, clinical, or safeguarding advice. Final responsibility for accuracy and appropriateness remains with the professional user.
