The AI divide: are we creating a new form of educational disadvantage?

Artificial intelligence (AI) could help close educational gaps, but unless we act deliberately, it could make them wider. For decades, education has grappled with a persistent challenge: children do not arrive at school with equal access to the resources, experiences and support that help them succeed. AI introduces a new dimension to that challenge.

A pupil with access to a high-quality AI tool, a knowledgeable parent and teachers who understand how to use AI may well benefit from explanations, feedback and learning support that extend far beyond the classroom. Another pupil may have limited access, little guidance and no reliable way of judging whether the answers they receive are accurate. Both may be described as having access to AI. Their educational opportunities are very different.

The next educational divide may not be between those who use AI and those who do not, but between those who know how to use it well and those who do not.

AI could change the experience of disadvantage

Consider a pupil who has struggled to understand a mathematical concept. At home, there may be nobody available who feels confident explaining it. An appropriately designed AI tool could offer another explanation, work through a similar example or help the pupil identify where their understanding has broken down. For a pupil with additional needs, AI may help make material more accessible. For a pupil learning English as an additional language, it may provide explanations that support access to the same ambitious curriculum as their peers.

For teachers, AI may create more time to identify misconceptions, adapt resources and respond to individual needs. These are significant possibilities. But they are not automatic outcomes. AI can also provide incorrect explanations, encourage dependence and make it easier to produce work without understanding it. The pupils who most need support may also be those least equipped to recognise these limitations. That is why simply providing access to technology is not enough.

An entitlement, not an advantage

Schools have long understood that literacy is more than access to books. Digital literacy is more than access to a device. Similarly, AI literacy must mean more than knowing how to enter a prompt. Every young person should learn how AI works at an appropriate level, what it can and cannot do, how to question its outputs, how to protect personal information and when independent thinking matters more than a convenient answer. They should also understand how AI is changing the world they will enter as adults.

This cannot depend on whether their school happens to have an enthusiastic member of staff leading AI development. AI literacy should become an entitlement for every pupil. That requires a corresponding entitlement for teachers: access to high-quality professional development, clear guidance and the time to develop their practice.

Equity must shape implementation from the beginning

The education sector has experienced previous waves of technology adoption in which access to devices was treated as the principal measure of progress. AI demands a more sophisticated response. School leaders should ask not only whether pupils can access AI, but whether its use improves learning for those who have historically benefited least from educational innovation.

Are disadvantaged pupils developing the knowledge to challenge AI-generated answers? Are pupils with SEND gaining greater access to ambitious learning? Are teachers receiving the support they need to use AI without lowering expectations? Are we evaluating these questions rather than assuming that innovation automatically produces improvement?

A national approach to AI in education should make equity one of its central tests of success.

The choice ahead

AI offers a genuine opportunity to make high-quality learning support more accessible. It could help teachers respond to needs that time and resources have previously made difficult to meet.

But without purposeful leadership, the benefits may flow disproportionately to those who already enjoy educational advantage.

At AiEd, we believe the ambition should be clear: the children who have the least should not be the last to benefit from AI. The question is not simply how quickly we adopt AI. It is whether we are prepared to ensure that its educational benefits reach every child.

‍