AI in Education: What School Leaders Need to Get Right

Artificial intelligence is already entering education. The challenge for leaders is ensuring that it strengthens learning, supports staff and prepares pupils for the future without compromising safeguarding, security, privacy or human judgement.

AI has moved remarkably quickly from something schools were discussing as a future possibility to something teachers and pupils can access today.

That creates an important challenge for headteachers, senior leadership teams, governors, trustees, safeguarding leads and IT leaders.

This is no longer simply a technology conversation. It is a leadership conversation involving teaching and learning, assessment, safeguarding, data protection, staff development, digital inclusion and the skills young people will need when they leave education.

Leadership Now Matters More Than the Technology

There are good reasons for schools to explore AI. Used appropriately, generative AI could provide useful support with areas such as lesson planning, educational resources, feedback, tailored support and administrative tasks.

But AI also brings important questions. AI-generated material can be inaccurate or inappropriate. Its use can create questions around personal information, pupil data, intellectual property, accountability and safeguarding.

That means successful AI adoption should not be measured by how many people are using it.

Education leaders should instead be asking whether AI is delivering a genuine educational or operational benefit while being used safely, appropriately and transparently.

---

1. We Need to Redefine AI Literacy

One of education’s most important responsibilities may become preparing young people to understand AI rather than simply teaching them how to operate AI tools.

It is relatively easy for a pupil to enter a question into an AI assistant. Real AI literacy involves knowing what to do with the answer.

Students increasingly need to ask:

• Can this information be independently verified?
• What might the AI system have misunderstood?
• Could the response contain bias or invented information?
• Where has the information come from?
• Is it appropriate to enter this information into an AI platform?
• Should AI actually be used for this particular task?
• Can I explain and defend the answer in my own words?

Young people need subject knowledge if they are going to recognise when an AI-generated answer is wrong. They also need critical thinking and digital judgement to use these systems responsibly.

In that sense, AI could make knowledge and critical thinking more important, not less.

---

2. Assessment and Academic Integrity Need a More Sophisticated Conversation

Much of the early debate around generative AI in education understandably focused on cheating.

The concern is legitimate, but concentrating entirely on detecting AI-generated work risks avoiding a more difficult question.

If an assignment is intended to demonstrate reasoning, subject knowledge or genuine understanding, schools need ways of assessing those qualities when pupils can access tools capable of generating sophisticated written responses.

This may require schools to develop much greater clarity around:

• when the use of AI is permitted;
• what acceptable AI assistance looks like;
• when AI use needs to be acknowledged;
• which tasks must be completed independently;
• how pupils demonstrate their reasoning and sources;
• how teachers establish that genuine learning has taken place.

---

3. Teacher Workload May Be One of AI’s Biggest Opportunities

There is understandably considerable interest in what AI can do for pupils. However, one of its most immediate benefits could be helping teachers and other education professionals with appropriate, repeatable tasks.

The most valuable AI implementation may not be a futuristic classroom filled with AI tutors.

It could be something far less dramatic.

If AI can appropriately reduce the time spent on certain administrative or preparatory tasks, allowing more time for teaching, mentoring, planning or direct interaction with pupils, then it deserves serious consideration.

---

4. Personalised Learning Is Exciting, but Evidence Must Lead

AI-powered tutoring and personalised learning attract considerable attention.

A system capable of adapting explanations, guidance and practice to an individual learner could offer interesting educational opportunities.

But personalisation should not automatically be equated with better education.

Before adopting a technology, education leaders should ask what evidence exists that it can improve the particular outcome their school wants to address.

Questions worth asking include:

• What educational problem is the technology intended to solve?
• What evidence supports its use?
• Has its effectiveness been independently evaluated?
• Is it appropriate for the relevant age group?
• How will teachers monitor its impact?
• Could its implementation unintentionally disadvantage certain pupils?

---

5. AI Must Not Come at the Expense of Human Relationships

Education is not simply the transfer of information.

Schools help young people develop confidence, resilience, curiosity, communication, collaboration and judgement.

Teachers also understand context. They build relationships, notice changes in behaviour and wellbeing and understand the individual circumstances surrounding a pupil.

AI may support teachers, but it should not reduce the value of professional judgement or the trusted relationships educators build with young people.

Technology should create more opportunities for valuable human interaction, not gradually engineer them out of the school day.

---

6. Could We Create an AI Capability Divide?

Historically, discussions about digital inclusion have often focused on whether pupils have access to suitable devices and connectivity.

AI introduces another dimension.

Two pupils might have identical laptops and internet access but completely different abilities to use AI effectively, critically and safely.

Differences in staff knowledge, pupil training and school-wide AI strategies could therefore become increasingly important.

This is why staff training and AI literacy deserve to be viewed as educational priorities rather than optional additions to an IT project.

---

7. Safeguarding, Privacy and Cyber Security Need to Be Involved from the Beginning

This is arguably one of the most important considerations for education leaders.

When a school begins using AI, questions about teaching and learning can quickly overlap with data protection, cyber security, safeguarding, intellectual property, system access and accountability.

A member of staff experimenting with a publicly available AI application is different from an institution formally adopting an approved AI system.

Before adopting AI more widely, leaders should consider:

These questions are not reasons to avoid AI. They are reasons to approach it professionally.

---

8. Governors and Trustees Need to Understand AI Too

AI cannot simply sit with the school’s IT team or IT provider.

It increasingly crosses curriculum, pastoral care, safeguarding, human resources, administration, assessment and information governance.

That means governors and academy trustees need sufficient understanding to provide appropriate strategic oversight.

They do not need to become AI engineers.

But they should be in a position to ask informed questions.

• Why are we adopting this technology?
• What problem are we trying to solve?
• What evidence supports the decision?
• What are the risks?
• What controls do we have in place?
• What training have staff received?
• How are pupils being educated about responsible AI use?
• How will we know whether the technology is delivering a genuine benefit?

---

Think Strategy Before Software

This may ultimately be the most important lesson for education leaders.

The first question should not be:

It should be:

For example, is the priority to:

• reduce appropriate administrative workload?
• support teachers?
• strengthen pupils’ AI and digital literacy?
• prepare students for future employment?
• support particular learning requirements?
• improve access to educational support?

Only once the problem and desired outcome are understood does the technology discussion become meaningful.

That is the difference between AI adoption and AI strategy.

Five Questions Every Education Leadership Team Should Be Discussing

These questions can help move the conversation from general interest in AI towards responsible decision-making.

---

AI Doesn’t Reduce the Importance of Education. It Changes Some of Its Responsibilities.

AI is capable of producing extraordinary things.

But perhaps its biggest impact on education will not ultimately be replacing something. It may be changing what we consider valuable.

When information is instantly available, evaluating information matters.

When AI can generate an answer, understanding whether that answer can be trusted matters.

When technology can create sophisticated content in seconds, critical thinking, originality, communication and human judgement could become even more valuable.