Search results

What School Leaders Need to Know About Ofsted's Approach to AI Artificial intelligence is increasingly present in schools, from pupils using ChatGPT for homework to administrative tools that help with attendance tracking. But how will Ofsted look at AI during inspections? In this blog, we share some of the key questions school leaders are asking. Will Ofsted specifically inspect our use of AI? No. Ofsted does not evaluate AI as a stand-alone element of inspections, and inspectors will not directly assess any AI tools you use. However, inspectors may consider the impact that AI use has on outcomes and experiences for children and learners. This includes both how your school uses AI and how you respond to AI use by pupils, parents and staff. Are we required to use AI in a particular way? No. Ofsted does not expect or require schools to use AI in any specific way, or to use it at all. At this stage, Ofsted acknowledges it doesn't have the evidence needed to define what constitutes "good use" of AI for inspection purposes. However, you should be aware that pupils and staff are likely already using AI in connection with education. Whether this is pupils using AI tools for homework or support staff using AI for administrative tasks, it's important that you're aware of where AI is being used, how it's being used and the impact this is having.  Will inspectors ask about AI during our inspection? Inspectors are not required to actively look for AI use, and inspection reports won't mention AI unless it's crucial to broader inspection decisions. Inspectors will only consider AI when it's relevant to their evaluations in areas like safeguarding, curriculum or attendance management. If inspectors find that your school's use of, or response to, AI has a significant impact on children and learners, they'll record it as evidence in the same way they would any other factor. What if pupils are using AI for homework or coursework? Where relevant, inspectors will explore how leaders ensure that AI use by pupils at school or at home is in their best interests. This might include looking at your school's policy on pupils using AI for homework and coursework. The key question is whether you've made sensible decisions about managing this use, just as you would with any other educational tool. What about the risks associated with AI? While Ofsted won't evaluate AI risks separately, these concerns will be addressed when they have implications for areas already considered during inspection: Data protection: Many AI applications use large amounts of data, including personal data. Inspectors may ask questions similar to those they ask about any data collection, storage and processing. Safeguarding: AI can pose unique safeguarding risks. Inspectors will consider this as part of their general evaluation of your safeguarding culture. For example, they may look at how you ensure pupils' use of AI on personal devices is in their best interests. Bias and discrimination: AI can perpetuate bias present in the data it processes. While this is a new context, inspectors already consider risks of bias and discrimination and the measures you've taken to mitigate them. Inspectors may ask what steps you've taken to ensure your use of AI properly considers these risks and whether you have assurance processes to identify emerging risks. Do inspectors need to understand how AI software works? No. Ofsted doesn't expect inspectors to understand how AI software has been designed, any more than they need to understand the programming of your firewall software to evaluate data security arrangements. The evaluation focuses on your decision-making, what you've considered, and the impacts on children and learners, not the technical details of the tool itself. Q: Can you give examples of how inspectors might consider AI? Here are some practical scenarios: If your school uses AI to identify causes of absence, inspectors may consider how it forms part of your overall approach to tackling absence. If you use AI-generated summaries of meetings or reports, inspectors may want to understand how you ensure these are accurate. If pupils are using AI inappropriately, inspectors may evaluate how you've responded and addressed the impact. If you're wondering how you could consider the use of AI in your setting, we share five areas you may wish to look into here. What if we don't use AI at all? That's absolutely fine. Many schools and pupils don't use AI, and Ofsted has no expectation that you should. The guidance simply clarifies how inspectors will approach AI if they do encounter it during an inspection. More information on how schools could approach AI can be found here. Key Takeaways for School Leaders ✓ You don't need to use AI, but be aware that pupils and staff probably are ✓ Make sensible decisions about AI use, just as you would with any other tool ✓ Consider the impact on pupils' outcomes and experiences ✓ Address the risks—particularly around data protection, safeguarding and bias ✓ Have policies in place for pupil use of AI in homework and coursework ✓ Document your decision-making about AI use and how you've considered risks Looking Ahead Ofsted has stated that this is not a final position and that their approach will be updated as understanding of AI use in education continues to develop. School leaders should stay informed but needn't feel pressured to radically change their approach. The same principles of sound decision-making and keeping pupils' best interests at the centre apply, whether you're using AI or not. This free editable download offers sample reflection questions across key areas of school life, plus prompts to help you think more widely about the ethical use of AI in education.

AI is already part of school life. You might not have a strategy yet or even a shared view across your staff team but someone’s probably tried it to plan a lesson. Someone else has tested it out for an administrative task (and maybe that person is you!) For many schools, it’s crept in quietly and that’s often how these things start. As a school leader, you may need to take a moment to think. What’s actually happening in your setting? Where are the risks? What might be worth exploring before it all beds in unnoticed? This post offers five thought-provoking prompts drawn from Honeyguide’s AI Use and Ethics Toolkit Pack – a practical, editable pack to help school leaders explore risks, responsibilities and leadership decisions across school life. 1. Pupil use of AI: Are we ignoring something that’s already happening? Most AI discussions in schools focus on staff but pupils are using it too, especially at home. They might be typing prompts into tools like ChatGPT to summarise a homework text or rewrite an answer in more ‘teacher-ish’ language – yes, even in primary.  But does this mean pupil use of AI is damaging? What about when they use it as a thinking partner to better understand tricky concepts or as a tool to promote independence? Here are some thoughts to consider: Is there a risk that our pupils will become over-reliant or use AI to shortcut thinking? Are we aware that AI tools can generate biased, inaccurate or age-inappropriate content, which, in turn, can affect pupils’ understanding? Could safe, structured AI use be built into lessons or homework to help our pupils use it effectively? How could we build pupils’ digital discernment and critical thinking, regardless of whether they currently use AI? 2. AI and SEND: Could it support personalisation or flatten the nuance? Some leaders are exploring whether AI could help generate first drafts of social stories, emotional scripts or visual schedules. In settings where pupils need clear, repeated routines — especially early years or specialist provision — this could save time and provide a useful starting point. But SEND support is rarely that straightforward. AI can’t know what tone feels right for an individual child, and there could be data protection implications if a staff member enters personal data about a specific pupil. Images created by AI may be too vague or abstract, and while generic scripts might look fine at first glance, they often miss the detail that makes a difference. Here are some questions to help you reflect: Could AI help reduce the time teachers or SENDCos spend creating basic scripts or visual routines? How do we ensure what’s produced still feels personal, safe and appropriate for each pupil? Might this help us respond more quickly when new needs emerge? What role do visuals and symbols play in our setting and can AI support this effectively? 3. AI and Parent Communication: Should we be upfront about its use? AI tools are already being used in some schools to help draft letters and emails to parents. Used carefully, this can support consistency and reduce admin time, especially when staff need to communicate clearly and professionally under pressure. But what happens when parents start to notice? If a message feels too generic or the tone slips, families may question how it was written. Some will be fine with AI involvement but others may see it as impersonal or even inappropriate, particularly in sensitive situations. Useful questions to consider: Should we tell parents if AI has supported part of a communication and in what circumstances? Would a clear internal stance help staff make good decisions here? How might different parent groups respond to this transparency? What safeguards do we have to make sure the school’s voice and values still come through? 4. Staff oversight: Do we risk being too hands-off or too heavy-handed? Most staff who are trying AI are doing so with good intentions — to save time, test ideas or ease workload. But without a shared conversation, small individual choices can quickly become whole-school habits with the potential for long-lasting damage. A good example of this is the use of AI to draft report comments: One teacher uses AI to help draft report comments to get started. Another writes every line manually, working late into the evenings and weekends. Another pastes in well-known generic banks of report statements and asks AI to make comments on each pupil based on a simple grading they give the pupil for effort and attainment. Another feeds AI with full pupil progress data and prompts it to create report comments for every pupil with positives, development points and targets. Over time, this kind of inconsistency creates problems, not just in tone or quality, but in fairness. Parents are reading these reports about their children and will likely expect them to feel personal, considered and accurate, even though staff are under huge pressure to get them written at all. If some are quietly relying on AI and others aren’t, it raises questions about workload, expectations and trust, as well as possible data protection implications. Some prompts to explore with your team: Do we know how and where staff are currently using AI and do they feel comfortable sharing it? Would some contexts benefit from clearer expectations or guidance (e.g. using AI to write pupil reports, using AI to plan lessons)? What support do staff need to use AI professionally without feeling monitored? How do we tap into the strengths of staff who are already exploring the use of AI? 5. Leadership and Governance: Who’s actually responsible? AI use in schools often begins quietly but when things go wrong, questions about oversight come quickly. Who approved that tool? Who checked that message? Who owns the decision? Without a clear line of accountability, schools risk leaving staff exposed and governors uninformed. If concerns are raised — whether by a parent, staff member or inspectors — ambiguity doesn’t hold up well under scrutiny. Questions to reflect on as a leadership team: Who currently oversees decisions about AI use across school operations? How are risks reviewed, shared and acted on when needed? Do governors understand the school’s current position and is it written down? Are we prepared to explain how decisions around AI are made and monitored? What can school leaders do to start a conversation about AI? You don’t need a full AI strategy to get started but you do need time and space to think. If these five prompts have sparked reflection, the AI Use and Ethics in Schools Toolkit Pack can help you go further. It includes seven editable checklists covering teaching, SEND, assessment, admin, leadership, ethics and more, with practical use cases, risks and questions to support confident decision-making.