Teachers are using AI for lesson plans and feedback. Students are using AI in their work — sometimes inside the rules, sometimes well outside. The institutions that navigate this well will have a competitive and reputational advantage.
The DfE released its generative AI guidance for education in 2023 and updated it in 2024. The position is measured and practical: AI can support teaching and learning, but institutions must have policies in place, staff must understand the risks, and safeguarding and data protection obligations are non-negotiable.
The gap between that guidance and what is actually happening in most schools, colleges, and universities is significant. Teachers are using AI tools to generate lesson plans, mark schemes, and feedback. Students are using AI in their work — sometimes within permitted boundaries and sometimes in ways that constitute academic misconduct.
The institutions that navigate this well will have a competitive and reputational advantage. The ones that don’t will find out the hard way.
AI can write essays, generate code, produce lab reports. Assessment frameworks built for a pre-AI world are not fit for purpose. The question is not simply “how do we detect AI use” — detection tools are imperfect and their use raises their own fairness concerns. It’s how institutions design assessments, communicate expectations, and handle suspected misconduct in a way that is consistent, fair, and legally defensible.
Children’s data carries enhanced protections under UK GDPR and the Children’s Code. AI tools used in educational settings — including LMS AI features, adaptive learning tools, and AI pastoral support tools — must meet these heightened standards. An EdTech vendor’s privacy policy is not a substitute for your own assessment.
Generative AI can produce images, text, and audio that create safeguarding concerns. Deepfake imagery involving students, AI-generated content targeting vulnerable young people, and the use of AI in bullying and harassment are emerging safeguarding issues. Governors, DSLs, and pastoral leads need frameworks to recognise and respond to these.
AI tools adapted for SEND students offer genuine accessibility benefits. They also raise questions about how AI modifications are documented in EHCPs, what the evidence base is for specific tools, and whether AI-assisted support is replacing rather than supplementing specialist human provision.
DfE guidance in practice. Ofsted’s evolving position on AI in schools. UK GDPR and the Children’s Code in educational contexts. Where exam board policies currently sit on AI use.
Building a position on AI in student work that is defensible, consistent, and age-appropriate. Assessment design principles for an AI-present world. Misconduct procedures that are fair, evidence-based, and legally sound.
EdTech AI tool assessment. Processing children’s data: obligations and red lines. DPA-compliant vendor selection process. Student data in AI tools — what staff need to understand.
Emerging safeguarding risks from generative AI. Deepfake content protocols. Online safety obligations in an AI context. Updating your safeguarding policy to reflect AI risks.
Governing how staff use AI in their professional practice — from lesson planning to marking to communication. What requires disclosure. Staff training approach by role.
| Use case | Efficiency gain | Primary risk |
|---|---|---|
| AI lesson planning tools | Significant time saving | Quality, curriculum alignment |
| Automated feedback and marking | Staff workload | Feedback authenticity, accuracy |
| Adaptive learning platforms | Differentiation | Data handling, SEND implications |
| AI writing assistants (student use) | Accessibility | Academic misconduct, equity |
| AI safeguarding monitoring tools | DSL capacity | False positives, data processing |
| Administrative AI (comms, scheduling) | Efficiency | Data handling, accuracy |
We assess where you stand against your sector’s regulatory floor and identify your highest-priority governance gaps.
One hour with leadership. Sector-specific regulatory framework, immediate priority actions, the language to take this to the wider team.
Half or full day. Full risk mapping for your sector, governance protocols, AI use policy draft, and 30 days advisory access. The flagship engagement.
Multi-site, network, and group pricing available on request.
Technology acceptable use policies were designed for internet access, device use, and online safety. They almost never address generative AI specifically — the circumstances of use, the academic integrity implications, or the safeguarding dimensions that AI content introduces. You need an AI-specific addendum at minimum, and ideally a standalone AI use framework.
AI detection tools have significant false positive and false negative rates. They may disproportionately flag non-native English speakers. Several institutions that relied on detection have faced legal challenges over unfair misconduct proceedings. Detection is one component of an academic integrity approach, not a substitute for a governance framework.
The answer is a clear, permissive policy that specifies approved tools, acceptable use cases, and disclosure requirements — combined with training so staff understand why the policy exists, not just what it says. Staff using AI to manage impossible workloads are not acting in bad faith. They need governance that acknowledges their context.
Tell us about your institution and your current AI situation — we’ll give you an honest assessment. No sales pressure. If your governance is sound, we’ll tell you.