Your Child's Homework Was Marked by AI. Did Anyone Think to Tell You?

AI is entering our classrooms. But are schools bringing parents, pupils and safeguarding arrangements along with it?

A friend recently told me about something that had happened at his son's school.

His son completed his homework, handed it in, and then received feedback. Nothing particularly unusual about that. Except the feedback included three words that caught his father's attention:

"Marked by AI."

My friend was surprised. The school hadn't told him it was using artificial intelligence to mark his son's work. His son hadn't been told either. As far as he knew, the school had not communicated how AI was being used, what safeguards were in place, or whether parents had any say.

His immediate reaction wasn't that AI should never be used in schools. It was much more straightforward.

Why hadn't anyone told him?

It's a perfectly reasonable question. In my view, it opens a much bigger conversation about how schools are adopting AI, who's overseeing its use, and whether the necessary governance arrangements are keeping pace.

The technology isn't necessarily the problem

I spend much of my professional life helping organisations understand and adopt AI responsibly. I advocate for what the technology can achieve, particularly when it frees people from repetitive tasks and lets them focus on work that genuinely benefits from human experience and judgement.

Education is an obvious example.

Teachers face considerable time demands. If AI can help prepare lessons, streamline administration, support assessment, and provide useful feedback, it could offer significant benefits.

More time for teaching. More individual attention for pupils. Less time spent on administration. Those are outcomes worth pursuing.

But there's an important distinction between using AI to support a teacher's professional judgement and allowing AI to make judgements without appropriate human oversight.

This is where my friend's experience becomes particularly interesting: did AI actually mark the homework, or did the teacher use it to draft feedback they later reviewed? Was the assessment checked for accuracy? What information about the pupil was shared with the AI system? Had the school assessed the risks associated with that particular use?

We simply don't know, and apparently neither did the parent.

That lack of understanding is the issue.

When does an efficiency become a governance concern?

Suppose an AI system marks a child's work incorrectly.

Perhaps it misunderstands the answer, overlooks a perfectly valid interpretation or provides feedback that is misleading, inappropriate or simply wrong.

Who is accountable?

The teacher? The school? The software provider?

What happens if the AI consistently disadvantages particular pupils, including those with special educational needs or disabilities?

What happens to the child's work once it's uploaded? Is it retained? Is it shared with a third party? Could it be used to train an AI model?

And what happens when a parent challenges the result?

None of these questions suggests that AI marking is inherently unsafe or inappropriate. Any school should be able to answer these questions before introducing a technology that influences a child's educational experience.

The important word here is before.

Governance shouldn't be something we retrofit after a concern has been raised.

Keeping Children Safe in Education 2026: AI is part of the safeguarding conversation

The September 2026 edition of Keeping Children Safe in Education (KCSIE) reinforces schools' and colleges' safeguarding responsibilities in England, including the need to recognise and respond to emerging technology-related risks.

The Department for Education's wider guidance on generative AI also emphasises transparency, data protection, human oversight, and the appropriate use of AI in educational settings.

These responsibilities need to be understood together. AI governance is not simply an IT matter, nor is it solely about preventing pupils from accessing inappropriate content online.

It extends to safeguarding, teaching practice, assessment, data protection, school leadership, and the trust relationship between schools and families.

The Designated Safeguarding Lead (DSL) has a particularly important role, but responsibility cannot reasonably sit with the DSL alone.

School leaders, governors, trustees, data protection professionals and teaching staff all have a part to play.

A school may have an established safeguarding policy, a data protection policy and arrangements for online safety. But have those arrangements been examined against how AI is actually being used now?

Do they cover teachers using AI to assess work? Staff entering pupil information into external tools? AI-generated communications to parents? Systems that offer personalised learning or make recommendations about a child's progress?

More importantly, can the school demonstrate that appropriate checks and balances are operating in practice?

Having a policy is one thing. Demonstrating that it works is quite another.

And that distinction matters enormously when we're talking about children.

Parents deserve to be part of the conversation

I suspect many parents reading this will recognise the situation.

We entrust our children to schools every day. We expect them to make professional decisions about how they teach, assess and support their pupils, but trust works best when it's accompanied by openness.

Parents don't necessarily need to approve every teaching tool a school uses. Equally, the absence of an individual consent form does not automatically mean a school has acted unlawfully. The requirements depend on how the technology is being used, what personal information is processed and the lawful basis for that processing.

Nevertheless, parents and pupils should understand how AI is used, especially when it processes their information or influences educational feedback and decisions.

If you're a parent, I would encourage you to ask your child's school a few straightforward questions:

Is AI being used to mark or assess my child's work, or to generate feedback?

Is a teacher reviewing the results before they're shared with pupils?

What personal information is entered into AI systems, and how is it protected?

How does the school check that the AI tools it uses are accurate, appropriate and fair?

Who oversees AI use, and how can parents raise concerns?

Where can I find the school's policy or guidance on how it uses AI?

These aren't hostile questions. They shouldn't be interpreted as a challenge to teachers' professionalism. They're reasonable questions about transparency, accountability and safeguarding and a school with sound governance arrangements should welcome the opportunity to explain them.

Helping schools demonstrate that their AI governance is working

That's why we developed AI Compass 4 Schools at Talisman AI Consulting.

Through our work in AI adoption, organisational readiness and human-centred change, we've seen that introducing a technology is often the easiest part of the process.

Ensuring people understand it, trust it, and use it responsibly is far more demanding.

Schools face a particular challenge because getting things wrong can have consequences that extend beyond operational inefficiency. They can affect children's safety, privacy, educational development and wellbeing.

AI Compass 4 Schools helps schools assess how effectively their existing arrangements address those risks.

It brings together a structured assessment of AI readiness, safeguarding and governance, aligned with relevant Department for Education guidance and the responsibilities reflected in KCSIE 2026.

It helps schools identify gaps in their arrangements, prioritise improvements and develop the evidence needed to demonstrate that AI is being introduced and managed responsibly.

Importantly, AI Compass 4 Schools is not intended to replace the DSL, the school's safeguarding framework or the professional judgement of its teachers and leaders.

Quite the opposite.

It's intended to support them by giving them greater visibility of the risks, clearer evidence of the controls already in place, and a practical way to identify what needs to improve.

For governors and trustees, it provides a more structured basis for challenge and assurance.

For school leaders, it helps connect AI adoption with the wider responsibilities of running a safe and effective school.

For DSLs, it helps ensure that emerging AI-related risks are considered within the school's established safeguarding arrangements, rather than treated as a separate technology initiative.

And for parents, the outcome should be greater confidence that schools are thinking carefully about the technology they're introducing into children's lives.

The real test of responsible AI

Returning to my friend's experience, I don't know whether the school had an excellent AI governance framework or no meaningful framework at all.

It would be unfair to draw conclusions from three words on a piece of homework, but I do know the parent didn't understand what happened, and that the school's use of AI came as a surprise. That alone is worth reflecting on.

Responsible AI adoption isn't just about having the right technology, choosing an approved supplier, or writing a policy.

It's about what happens in practice. Whether staff understand their responsibilities. Whether safeguarding risks have been considered. Whether decisions can be explained and challenged. And whether those affected by those decisions have enough information to trust the process.

Schools should absolutely be exploring the opportunities AI offers. I would argue that, used well, it can make a meaningful difference for both teachers and pupils.

But enthusiasm for the technology cannot outrun the responsibilities that come with using it.

The question isn't whether AI has a place in our schools. It does.

The question is whether schools can demonstrate that its use is safe, proportionate, transparent and properly governed.

Every school leader should be able to answer that question; equally, every parent should feel comfortable asking.

About AI Compass 4 Schools

AI Compass 4 Schools is a dedicated AI readiness, governance and safeguarding support framework developed by Talisman AI Consulting and Pro Apprenticeships.

It helps schools and multi-academy trusts assess their AI arrangements against relevant Department for Education guidance, identify governance and safeguarding gaps, and develop practical, evidence-based improvements. Importantly, it sits alongside and in support of the DSL role.

Designed to complement existing safeguarding responsibilities, it supports DSLs, senior leaders, governors and trustees in understanding and overseeing the responsible use of AI.

Find out more about AI Compass 4 Schools at Talisman AI Consulting.

Roger Wilson Founder, Talisman AI Consulting Where AI meets human-centred change.