The short answer
AI in schools is most useful when a school starts with a specific educational or administrative need, then checks whether a tool is safe and suitable for that task. Generative AI can help staff draft or adapt materials, but its output needs human review. Evidence about pupils using generative AI themselves is still developing, so student-facing use needs especially clear boundaries and oversight.
Where AI may help in a school
The DfE describes potential uses such as drafting lesson resources, questions, quizzes, revision activities and communications. These are starting points for staff to check and adapt, not finished materials to use without review.
- Planning: generate a first draft of a lesson activity or discussion prompt for a teacher to check.
- Practice: create question variations or revision prompts that staff can compare with the taught curriculum.
- Administration: draft routine communications or summaries, with a member of staff checking accuracy and tone.
- Guided learning: explore bounded, curriculum-led support where the school has assessed the tool, the content and the oversight arrangements.
A tool that saves time in one context may add checking work in another. Define what success means for the task, and evaluate quality and workload rather than assuming that using AI is itself an improvement.
What schools should assess before use
The DfE’s safe-use materials highlight safeguarding, ethics, data and intellectual property. A school’s review should also consider whether the tool is reliable, accessible and appropriate for the age group and intended use.
A practical evaluation checklist
Before a school approves an AI tool or a new use of one, decision-makers can ask:
- What problem are we trying to solve, and how will we know if the tool helps?
- Is the proposed use permitted by our policies and appropriate for the pupils and staff involved?
- What pupil or staff data is involved, and has our data protection lead reviewed the processing?
- What happens when the AI is wrong, biased, inappropriate or unavailable?
- Who checks outputs, responds to concerns and decides whether the use should continue?
- Can we run a limited pilot, gather feedback and stop or change the use if it does not meet our criteria?
Ask the supplier for information about the exact deployment, data flows, retention, subprocessors, access controls and safeguarding arrangements. Assess the complete service rather than relying on a product label or a claim about one part of the system.
A measured way to get started
- Choose one defined use. Keep the first test narrow enough to review properly.
- Agree the boundaries. Identify approved users, materials, data, staff responsibilities and escalation routes.
- Brief the people involved. Explain what the tool can and cannot do, and how to raise concerns.
- Review the evidence. Compare the pilot with the school’s intended outcomes, including accuracy, learning value, workload and pupil experience.
- Decide what happens next. Continue, adjust or stop based on the review, and update school guidance when needed.
The DfE recommends that AI use is carefully planned and that what is appropriate should be judged in the context of each setting. Its free support materials can help staff understand risks and mitigations.
Official guidance and further reading
- DfE: Generative artificial intelligence in education Department for Education policy paper on opportunities, risks and the safe and effective use of generative AI. Applies to England.
- DfE: Safe use of generative AI in education — module 3 Staff training on safeguarding, ethics, data, intellectual property and risk mitigation. Applies to England.
- DfE: Generative AI and data protection in schools Guidance on personal data and generative AI in school settings.
- DfE: Using AI in education settings — support materials Free resources for education staff on safe and effective use.
Source links checked 2 October 2026. Always refer to the latest official guidance and your school’s own policies.