AI Ethics in Education: What Teachers Need to Know [2026]

AI tools raise ethical questions that most teachers have never needed to consider. When you use ChatGPT to draft a worksheet, who owns the output? When a learner submits work through an AI marking platform, where does that data go? When an adaptive learning system decides a learner needs easier material, is it helping or labelling? These are not abstract philosophical problems. They are decisions that UK teachers face every week, and the answers shape what kind of education learners receive.

The current DfE guidance on generative AI in education asks schools to use AI safely, effectively and responsibly, with professional judgement, data protection and safeguarding built into any use. This article turns that guidance into practical actions for teachers and leaders.

‍

The Five Ethical Dimensions

AI ethics in education is not a single topic. It covers five distinct areas, each with different implications for how you use AI in your classroom and your school.

Dimension	Core Question	School Responsibility
Privacy	Where does learner data go?	Data Protection Impact Assessment for every AI tool
Bias	Does the AI treat all learners fairly?	Regular audit of AI outputs across demographic groups
Transparency	Do learners and parents know AI is being used?	Clear communication in school policy and parent letters
Autonomy	Are learners developing thinking skills or outsourcing them?	Curriculum design that builds metacognitive independence
Accountability	Who is responsible when AI gets it wrong?	Teacher remains the accountable professional in all cases

Learner outcomes rely on transparent and effective AI use. Opaque tools make accountability hard, (O'Neil, 2016). Clear but biased AI creates unfairness, (Noble, 2018). Focus on all five ethical AI areas, (Holmes et al., 2021; Selwyn, 2022; Luckin, 2023).

‍

Data Privacy: The Non-Negotiable

Data privacy is the most immediate ethical concern because it has legal force. UK GDPR applies to all processing of learner data, and schools are the data controllers responsible for compliance. When a teacher pastes learner work into ChatGPT, that is a data transfer. When a school deploys an adaptive learning platform, that is data processing at scale.

The practical requirements are specific:

Data Protection Impact Assessment (DPIA). Required before deploying any AI tool that processes learner data. The DPIA must document what data is collected, where it is stored, how long it is retained, and whether it is used for purposes beyond the school's intention (such as model training).

If a tool processes learner data, schools need to understand where that data is processed, stored and shared, and whether the arrangement complies with UK GDPR and school policy. Avoid naming any provider as "UK-safe" unless its privacy notice, data-processing agreement and current data-residency terms have been checked by the school's DPO or IT lead.

Some AI tools use submitted content to train or improve their systems. Before entering learner work, staff should check the approved tool list, vendor terms and data-processing agreement. Where pupil work or personal data could be used for model training, do not use the tool unless the school has a lawful basis, appropriate permissions and clear DPO approval.

Upload only essential data to the tool. Remove learner names, school details, and personal information. Use candidate numbers or initials instead. This lowers data protection risks, and the tool still gives helpful results.

DfE data-protection guidance for schools says staff should check that an AI tool is approved by the school, understand how it uses personal data, acknowledge or reference generative AI use, and fact-check outputs. Keep an internal list of approved tools, owners and permitted use cases, then review it when tool terms or school policy changes.

‍

Bias in AI Education Tools

According to O'Neil (2016), AI learns from biased data. This data often mirrors current inequalities. UK teachers must watch for three specific effects in education. Buolamwini and Gebru (2018) and Noble (2018) show how bias affects learners.

AI writing tools favour standard English. Learners using dialects or idioms may score lower (Bridgeman et al., 2012). This reflects language, not ability. UK dialect research is limited, but the risk remains.

Automated scoring can be sensitive to surface features such as length, structure and standard-language patterns, so teachers should treat AI marking as a first-pass support rather than a final judgement. Shorter, precise answers, regional language patterns or concise methods such as PEEL may need human review, especially for SEND, EAL and dialect variation.

Adaptive platforms may limit learner progress (Francis et al., 2020). Algorithms offering only easier content prevent learners facing needed challenge. This mirrors issues with rigid setting in schools. AI should support access to complex content.

Use AI tools with bias checks. Review AI marks and feedback for all learner groups. Compare AI assessments to your own judgement. If patterns appear (e.g., AI underscoring EAL learners), flag it and change your method.

‍

Transparency with Learners and Parents

Learners and parents have a right to know when AI is being used in assessment and teaching. This is both an ethical principle and a practical one: trust in assessment depends on understanding how it works.

What to communicate:

A straightforward approach is a brief section in the school's assessment policy stating: "This school uses AI tools to support marking and resource preparation. AI-generated feedback is always reviewed by a teacher before being shared with learners. No AI tool is used as the sole basis for any grade that contributes to reporting." This sets expectations without creating unnecessary concern.

Learner understanding:

Learners benefit from understanding AI's strengths and weaknesses. Year 9 learners can compare an AI comment with their own reasoning and ask, "Was my method right even if the tool marked it wrong?" This kind of review builds metacognition and AI literacy because learners evaluate feedback rather than accepting it automatically (Long & Magerko, 2020).

Parental communication:

Parents do not need a technical briefing on AI architecture. They need reassurance that: (1) their child's data is handled securely, (2) AI supplements rather than replaces teacher judgement, and (3) the school has a policy governing AI use. A paragraph in the school newsletter or a dedicated section on the school website meets this need.

‍

Learner Autonomy and Thinking Skills

AI's ethics concern teachers: does it help learners think or hinder them? If AI plans essays and corrects spelling, what thinking do learners do? (Holmes et al., 2023).

Bjork (1994) found that struggle helps learning. Learners build stronger memories when they face problems. Working through confusion creates deeper understanding. AI tools removing learning friction might reduce retention.

The practical question is where to draw the line. Using AI to generate a first draft teaches learners nothing about writing. Using AI to provide feedback on a draft they wrote themselves builds their capacity for revision. Using AI to check factual claims teaches critical thinking. Using AI to generate the facts teaches recall without understanding.

AI Use	Builds Thinking	Undermines Thinking
AI provides feedback on learner's own work	Yes: learner evaluates and responds to feedback
AI generates the first draft		Yes: learner skips the cognitive work of composition
AI generates retrieval practice questions	Yes: learner engages with recall and spacing
AI summarises a textbook chapter		Yes: learner bypasses comprehension and selection
AI offers scaffolding hints during problem-solving	Yes: learner still does the reasoning with support
AI fact-checks a learner's claims	Yes: learner develops source evaluation skills

AI should support learning, not replace it. The DfE guidance is clear that generative AI can make some written tasks quicker, but it cannot replace human expert judgement, subject knowledge or responsibility for final content. Use AI for low-risk drafting and formatting; keep high-value learning tasks, argument building and creative expression with learners.

‍

Building a School AI Policy

DfE guidance says schools and colleges may wish to develop guidance on when it is acceptable or appropriate to use generative AI and how they engage parents. Many schools turn this into a short written policy or appendix that answers six practical questions.

Policy Question	What to Include
Which AI tools are approved?	Named list of tools reviewed by DPO, with approved use cases for each
What data can be shared with AI tools?	No learner names or identifiable data. Anonymised work only, via approved tools.
How is AI used in assessment?	AI for formative assessment only. All grades reviewed by teacher before recording.
What are learners allowed to do with AI?	Clear rules by key stage. See academic integrity guidelines.
Who reviews the policy?	Named lead (often the computing lead or a deputy head), annual review cycle
What training do staff receive?	Minimum CPD requirement before using AI tools. Ongoing updates as tools evolve.

A one-page policy that answers these six questions clearly is more useful than a 20-page document that no one reads. The goal is a shared understanding across the school, not a compliance exercise. For guidance on creating your policy, see our guide to creating an AI policy for schools.

‍

AI Ethics in the Classroom

UNESCO guidance on generative AI in education and research frames AI education around human-centred, ethical, safe and equitable use. Computing and PSHE offer suitable spaces, yet the concepts apply widely. Teachers can help learners in all subjects ask how AI systems use data, make errors and affect people.

KS2 (Ages 7-11): Introduce the concept that AI tools can make mistakes and that people need to check AI outputs. A Year 5 class can evaluate AI-generated text for factual errors, building both critical thinking and AI awareness. Frame it as: "The AI is a helpful tool, but it does not always get things right. Your job is to check."

KS3 (Ages 11-14): Explore bias in AI systems. A Year 8 class can test whether an AI writing tool gives different scores to the same content written in different styles or dialects. This teaches both AI literacy and awareness of systemic bias. Link to citizenship and critical thinking curricula.

KS4 learners can study AI ethics and its societal impact. Learners in Year 10 can analyse AI marking tool data pipelines. They can explore data use, decision-making, and bias potential. This links to "ethical, legal... impacts" (DfE National Curriculum).

These lessons develop learners' AI skills for life after school. They also enable learners to use AI tools critically, rather than accepting outputs without thought (Long & Magerko, 2020). This fits the DfE's expectation that pupils learn about the limitations, reliability and potential bias of generative AI.

‍

Environmental Considerations

Strubell et al. (2019) found training AI uses much power, causing emissions like flights. Single AI tasks seem small, but they add up. AI's overall impact on learners' education is real.

For schools, proportionality is key. Use AI when it adds educational or workload value, such as drafting quiz items for teacher review, and avoid using it for basic tasks that are quicker without it. Strubell et al. (2019) and Patterson et al. (2021) support the wider point that model development has environmental costs, but they do not prove that every school AI task saves carbon.

‍

Accountability: When AI Gets It Wrong

AI tools will make errors. An AI marking tool may misgrade an essay. An adaptive platform may route a learner to inappropriate content. A chatbot may provide factually incorrect information. When this happens, the accountability sits with the school and the teacher, not with the technology vendor.

The DfE guidance says AI-produced content requires professional judgement and that the quality and content of the final document remain the responsibility of the person and organisation using it. Treat AI output like an untrusted draft: check accuracy, bias, context and safeguarding before it reaches learners.

Kasneci et al. (2023) discuss a human-in-the-loop approach to large language models in education. In school practice, that means teachers check AI work before learners use it, verify AI-generated feedback before it enters markbooks, and make sure adaptive pathways remain suitable for each learner. AI may be faster, but teachers retain responsibility for quality.

‍

Starting Points for Your School

Implementing ethical AI use does not require a complete overhaul. Start with three concrete steps that any school can take within a term.

Step 1: Audit current AI use. Survey staff to identify which AI tools are already being used, how they are being used, and whether they have been reviewed by the DPO. Many teachers are already using ChatGPT or similar tools informally. The audit brings this into the open so it can be governed properly.

Step 2: Write the one-page policy. Using the six questions above, draft a policy that covers approved tools, data handling, and assessment use. Share it with all staff and include it in the staff handbook. Review it annually.

Step 3: Run one CPD session. A 30-minute session covering what AI tools can and cannot do, the school's approved tool list, and the data protection requirements. This does not need to be a full training day. A focussed, practical session during a staff meeting is sufficient to establish a baseline of understanding.

For a broader overview of AI tools and their classroom applications, see our hub guide to AI for teachers. For assessment-specific guidance, see AI and student assessment. And for the related question of how learners use AI in their own work, see our guide to AI and academic integrity.

‍

UK Regulatory Landscape for AI in Schools

UK schools follow data-protection, safeguarding, assessment and online-safety duties when adopting AI tools. Current DfE and Ofsted guidance asks schools to consider data protection, safeguarding, bias, discrimination and the impact of AI on children and learners. The ICO's Children's code is also relevant where services process children's personal data.

‍

4 Key Regulatory Requirements

(1) Data Protection Impact Assessment (DPIA) before adoption. Before your school uses any AI tool that processes learner data, you must complete a DPIA. This is a legal requirement under UK GDPR Article 35. The DPIA asks: what data does the tool collect? Where is it stored? Who has access? How is it protected? Could it be misused? A secondary school is considering using an AI marking tool. The DPO completes a DPIA: the tool collects student answers to essay questions (sensitive educational data). The vendor is US-based and subject to US government data requests. The DPIA flags: "Medium risk, US location, but vendor has EU data centre option." The school negotiates with the vendor to use the EU data centre, reducing risk to low.

(2) Clear information to parents about AI use. Parents must be informed that AI is used in their child's education. This is a transparency requirement under UK GDPR Article 13 and 14. Your privacy notice must explicitly mention AI. A primary school uses AI to generate differentiated worksheets for reading. They update their privacy notice: "We use generative AI (ChatGPT) to create personalised learning materials. Student names and work are not uploaded; we only input learning objectives and learner needs. The AI output is reviewed by staff before use." They send this to parents in the newsletter.

(3) Human oversight of AI output used for assessment or reporting. If you use AI to mark work or write student reports, a human teacher must verify the output before it goes to students or parents. AI alone should not make assessment decisions that affect learners. The ICO gives specific guidance on profiling children and automated decisions, and DfE guidance expects schools to be open and transparent where automated decision-making or profiling is being considered.

(4) Compliance with UK GDPR and the Children's Code. The Children's Code requires online services likely to be accessed by children to put children's best interests, privacy and transparency at the centre of design. If a free public AI tool does not provide age-appropriate safeguards, clear data-use terms or school-level controls, do not approve it for pupil use. Staff can still use approved tools for planning when no pupil data is entered and outputs are checked.

‍

Practical Checklist for School Leaders

Before approving any new AI tool, use this checklist:

1. DPIA completed. Has the DPO assessed risk? Are there unacceptable risks?
2. Privacy notice updated. Do parents know we use this tool?
3. Vendor agreement in place. Does the vendor have a DPA or signed agreement confirming GDPR compliance?
4. Data location verified. Is learner data stored in the UK or EU (preferred) or US (higher risk)?
5. Human oversight defined. Who verifies AI output before it affects learners?
6. Children's Code compliance. Does the tool meet standards for child safety?
7. Staff training provided. Do staff know how to use this tool safely?
8. Opt-out available. Can parents opt their child out if they wish?

‍

Classroom Example: Data Protection Officer's Review

A data protection officer review should focus on the tool's purpose, data categories, data location, model-training terms, processor agreement, age restrictions, safeguarding features, accessibility, bias risks and human oversight. The question is not whether AI tools can help learners in general; it is whether this specific tool is safe, lawful and educationally useful for this specific use.

Tool A (UK vendor, EU data storage): Passes DPIA, Privacy Notice updated, no concerns.

Tool B (US vendor, US data storage): DPIA flags medium risk. Vendor has UK subsidiary and can offer EU data centre. DPO negotiates contract. Risk acceptable if EU centre is used.

Tool C (unclear data use): DPIA fails because the vendor uses learner work for model training, data location is unclear and no data-processing agreement is available. Reject the tool for learner data until the vendor can provide acceptable contractual, privacy and safeguarding evidence.

This due diligence protects the school from regulatory action and ensures learners are protected.

References:

• Department for Education. Generative artificial intelligence (AI) in education.
• Department for Education. Generative artificial intelligence (AI) and data protection in schools.
• Information Commissioner's Office. Introduction to the Children's code.
• Information Commissioner's Office. Children's code standards.

Link: Creating an AI Policy for Schools 2025

‍

Frequently Asked Questions

‍

‍

‍

Further Reading