VTB experts named the basic principles of teaching children how to work safely with AI

On the eve of Children's Day, VTB experts named key rules that will help parents teach their children how to work safely with artificial intelligence (AI).

According to experts, the main thing is to explain to children the nature of technology and build the right boundaries for its use.

"Artificial intelligence is not a living expert or an ordinary encyclopedia, but a complex tool that can be both correct and wrong. A child who understands this principle is already protected: he will not perceive the AI's answers as unconditionally correct and is less likely to trust him with something that requires human involvement. Therefore, it is worth starting a conversation about neural networks not with intimidation or a list of prohibitions, but with an explanation of how they actually work," said Alexey Pustynnikov, leader of the VTB model development team.

The bank's experts recommend starting AI training with creative tools, such as image generation services. It is better to connect text neural networks later, when the child has already learned how to evaluate information and double-check facts.

Experts also advise parents to use AI with their child: doubt the chatbot's answers out loud and check them. It teaches critical thinking more effectively than any instructions.

Also among the recommendations is to treat AI as a simulator for solving problems. The right question for a neural network is not "decide for me," but "explain how to come to a decision and why it's the right one." This approach develops thinking, not replaces it.

"AI is perhaps the best tutor possible: it never gets tired, explains the same thing as many times as needed, and adapts to the student's level. The child may ask him to sort out a difficult topic step by step, practice answering questions before the test, or check his own solution. The main thing is to remember that the final word always remains with the person," added VTB expert Anna Shirshova.