Should the use of Gen AI be banned among younger students in schools?

Recently, New York City Mayor Zohran Mamdani announced a one-year moratorium on most generative AI tools for public school students until Class 8 for the 2026-27 academic year. Officials argued that young children require human connection and independent problem-solving skills instead of relying on algorithms.

India’s policies on this issue are still evolving. So, should the use of Gen AI be banned among younger students in schools? Lavina Jaswani and Deepa Jacob discuss this question in a conversation moderated by Priscilla Jebaraj.

What kind of AI tools do you use and how do they benefit primary school children? Lavina Jaswani: We have been using a number of AI-enabled tools to explain concepts to children; to offer personalised learning; and to help teachers create resources, from stories and worksheets to different types of assessments and quizzes.

There are different apps we use, such as Kahoot! and Quizlets, to offer a gamified, interactive, and sensorial learning experience that keeps children engaged, rather than the traditional rote learning method. However, the teacher still has to ask herself: what is my learning objective?

What does a child need to understand? Where can the technology add value and is it truly beneficial and essential for the child? Keeping such questions in mind, AI tools can be effectively used in the classroom. Deepa Jacob: In our school, technology is integrated seamlessly into our teaching and learning experience.

Every child from kindergarten onwards has access to an iPad, which is kept at the school; they don’t take them home. They have apps that might be able to assist them with learning such as Seesaw, which is a journal of their learning through which they can keep a record of things that they do, like published writing pieces, videos of science experiments, their reflection on their learning etc.

You can even assign tasks on them. Also read | Why is New York City pausing some AI tools in schools? Very recently, I discovered that Seesaw itself has an AI component. So, for example, we do a lot of work with reading fluency.

The kids are given a grade-level passage. They do a cold read of it on day one and through the week they reread the passage along with the teacher, and learn to pronounce the words, the meaning of the words, and then finally on day five they record their reading again and they find that they’ve improved in their fluency and their accuracy.

Recently, one of my colleagues found out that the recording feature has an AI component which transcribes what you’ve said and scores you based on your accuracy. This is really great because rather than the teacher having to listen to every child read, if an app can score your reading and tell you which words you are reading incorrectly, then you can practise those words alone.

But it’s AI, not a person, so it comes with its own flaws. For example, one of my students was reading a word with an Indian accent, and the app marked it wrong. The same thing happened to students using several other Asian accents.

So, that’s where I feel the human component is important. We need to be careful that we don’t let AI take over our lives. I also do not allow generative AI in my classroom. When I ask my children a question, or if they have a question, I will never ask them to Google it.

Because the easiest way to kill a love for learning is to find a ready-made answer. Also read | Five must-have AI tools for students Educationists and philosophers in the field of education have talked about existing in the liminal space, which is that space where you struggle with a concept, that threshold between not knowing something and struggling with something and then finally learning it.

I think that is a beautiful space to be in mentally and we need these as human beings. We need this time and space to reconfigure ourselves; time for ideas to incubate and grow. Relying on technology in that phase would be terrible.

While technology should be used as a tool, it is easy to over-rely on technology and use it as a crutch. If you want to create human beings that make their own contribution in the world, I think there should be serious restrictions on the amount of AI used among kids.

One fear is that AI could harm a student’s ability to think and learn for themselves — for instance, if they just enter their homework as a ChatGPT prompt, the AI does their homework for them. Is this a valid fear? Lavina Jaswani: Yes, this is where we need to take a balanced position.

AI can affect children’s thinking abilities, but it depends on how we use it. There are two possibilities: one is when the child uses the AI, the AI gives the answer, the child copies it, submits it and moves on. Then we have a problem, as the child is not using his brain.

Instead, we need to guide the child on how to use it and that is where we need to put the restriction. Also read | Why students need to read before they use AI Suppose you are trying to teach fractions or some kind of problem-solving lesson.

If the AI gives the answer, the child misses the struggle, the trial and error, the logical thinking, and the satisfaction of finally getting the answer. So the rule in my classroom is that first, the child needs to think it, then write it and then lastly, ask the AI to verify the answer.

They must try to solve it themselves, but if they can’t, just as they may ask their parents or their teacher for help, they can ask AI too to provide assistance in the same way. Podcast |Is Artificial Intelligence affecting critical thinking skills?

Deepa Jacob: The problem is over-reliance. If you’ve been driving a manual car and then you shift to an automatic car, you’re going to find it really hard to go back to driving manual, right? Similarly, I think you can become over-reliant on someone else doing the thinking for you or even synthesising ideas by just feeding them into an AI tool such as the Notebook LLM.

The capacity of the human brain is immeasurable. There’s...