AI could mean social mobility...or it could widen gaps

AI could help boost achievement among disadvantaged pupils. But, says Tom Moule, we need to develop a shared vision of ethical practice

Tom Moule

Human and robot hands touching

There’s a quote from Malcolm X that really resonates with me: “Education is the passport to the future, for tomorrow belongs to those who prepare for it today.” 

As a teacher, it was my responsibility to ensure that students developed into well-rounded individuals, ready to face life’s challenges and lead fulfilling lives. Naturally, Malcolm X’s words hit home here. 

Out of the classroom, my work now focuses on the ethical implications of artificial intelligence in education – and these words have taken on a new life. 

With the introduction of artificial intelligence (AI) into education, some of our most cherished educational goals hang in the balance. For instance, AI could help bring about a paradigm shift in social mobility, but it could also lead to a two-tier system, with staggering educational inequality across the digital divide


We’re at a crossroads. On one side is a vision of compassion, opportunity and optimism. On the other lies intensive stratification. There can be no doubt about which version of tomorrow we should be preparing for today. 

We must prepare for a more socially mobile society than we currently have. Research from the Education Policy Institute has shown that on average, disadvantaged pupils leave school the equivalent of 18 months behind non-disadvantaged pupils. And this gap is widening. 

Educators are already using AI systems to enhance learning through pinpointed personalisation and increased engagement, allowing learners’ unique needs and strengths to be catered for. If the same world-class innovations are available to all, life chances will truly be levelled up. 

AI also presents a credible opportunity for affordable, flexible and high-quality lifelong learning, allowing people to upskill with ease throughout life. 

A steer in the right direction

AI may not be a silver bullet for social mobility, but it could steer us decisively in the right direction. It could also do the opposite. 

AI systems are prone to reinforcing biased patterns in their training data, meaning they could force disadvantaged learners down less aspirational paths. Inequalities in access to devices could contribute to widening disparities in provision. 

And, even if we succeed in implementing unbiased, accessible and highly effective AI systems across the board, there is the danger of overuse. Governments across the world may decide that AI’s efficacy justifies deskilling teachers and switching to a cheaper workforce, because AI plus human chaperoning leads to a viable level of learning. 

And, even though viable may have to do for the many, the privileged few are likely to demand much more, such as a blended approach where excellent teachers – freed from cumbersome admin – use augmentative AI tools to enhance their practice. 

If AI makes excellence the norm, social mobility will increase; if AI merely makes the norm cheaper, then social mobility will suffer.

Preparing for a brighter tomorrow

Social mobility is not the only issue that hangs in the balance. Diversity and gender equality are also in a precarious state. If algorithmic bias creeps in, it could exacerbate existing asymmetries. Then again, AI can provide deep insights for teachers and leaders, which could facilitate smart strategies to level the playing field.

In preparing for a brighter tomorrow, we must safeguard learners and wider society from the potential downsides of artificial intelligence. But we must also facilitate effective, responsible use of AI so that we can realise a more diverse, socially mobile future. 

The first step is to unite those with a stake in AI in education – school pupils, college and university students, adult learners, teachers, lecturers, policymakers – and develop a shared vision of ethical AI, which reflects the views and values of those who have so much to gain, yet so much to lose. 

The second step is to establish the structures needed to support that vision, which may range from regulation and standards to awareness campaigns and training. 

We are confident that these preparations can serve as our passport to the desired destination. But to complete this journey, collective determination will be necessary. Everyone can play their part by demanding the best possible deal from AI in education.

Tom Moule is the executive lead at the Institute for Ethical AI in Education

Register to continue reading for free

It only takes a moment and you'll get access to more news, plus courses, jobs and teaching resources tailored to you

Latest stories

The link between language development and behaviour in schools

Creating behaviour policies in multicultural settings

The array of cultural backgrounds of people who meet and mingle in international schools can make creating behaviour policies that everyone can follow tough – but it has to be done. Dan Worth finds out how
Dan Worth 21 Sep 2021