AI tutors could become available to every child while access to a human teacher becomes increasingly limited to families and schools able to afford it, an education expert has warned.
The warning comes as schools face teacher shortages, pressure on budgets and growing investment in artificial intelligence. The author argues that technology may widen, rather than close, the gap between children unless it is used to support teachers rather than replace them.
Children from lower-income households are already using chatbots more heavily than their wealthier peers. Pew Research Centre data cited in the commentary found that one in five teenagers in families earning less than 30,000 US dollars a year do all or almost all of their homework with chatbots, compared with 7% of those from households earning 75,000 dollars or more.
But access to information does not necessarily lead to learning, the author argues. Without an adult to guide and encourage them, children may become dependent on automated help and lose the ability to work through difficult problems independently.
Why AI tutors cannot replace human teachers
An experiment involving about 1,000 Turkish high school pupils examined the effect of using standard ChatGPT, a tutoring version of the chatbot, or no AI while solving maths problems.
Both groups using the chatbots performed better during practice. However, when they sat an examination without the tools, pupils who had used standard ChatGPT performed 17% worse than those who had worked without AI.
The findings suggested that some students had become reliant on hints rather than learning how to solve the problems themselves. The author says pupils are willing to hand difficult tasks such as writing an essay to AI, even though they would not similarly outsource activities they find enjoyable.
The writer, who has a PhD in education and comes from a family of educators, argues that teachers have a role that technology cannot easily reproduce: making learning feel worthwhile. Motivation, they say, is more likely to come from a person who understands a child than from an algorithm.
They recall a young man who arrived at an IT school about 15 years ago looking as though he was homeless. After being told he wanted to study, he raised the money for the course, completed it and later taught at the school.
The man later said he had succeeded because teachers believed in him and encouraged him not to give up. An AI tutor could have provided him with information, the author says, but it would not necessarily have given him the motivation to acquire the knowledge.
Teacher shortages increase pressure to use AI
The United States Department of Education has included AI development among its priorities for federal school grants, specifically mentioning AI tutors and support for pupils who are falling behind.
During consultation on the plans, commentators warned that classroom AI could reduce the time children spend interacting with teachers. The department said the human element of education should be preserved, but the priority remained unchanged.
The technology is advancing rapidly and becoming cheaper, while qualified teachers remain a scarce and costly resource. UNESCO forecasts that an additional 44 million teachers will be needed around the world by 2030, including 4.8 million in Europe and North America. In the United States, 56,000 teaching positions are reportedly unfilled.
Teachers’ salaries account for nearly 30% of spending in American public schools, making them the largest expense for school districts. Faced with shortages and possible budget cuts, some districts are already considering cheaper alternatives to employing more teachers.
The author warns that this could reduce teachers to platform operators who monitor performance data and intervene only when an AI system identifies a problem. Schools must instead decide whether AI will give teachers more time to support pupils or be used to justify employing fewer of them.
That choice, the commentary argues, will determine whether children continue to have a teacher who believes in their potential, or whether a technology intended to level the playing field ends up reinforcing educational inequality.
