What research tells us about the role of artificial intelligence in the life of a child

Artificial intelligence, also known as ‘AI’, refers to the nifty techniques used by machines, software, and technology to solve problems without human help. AI is becoming increasingly common in our day-to-day life and in that of children. A well-known example of AI is the smart algorithm Netflix and Youtube use to offer their users, and with that children, a personalized range of series and videos. But AI can also be found in smart toys, such as Barbie dolls that can verbally respond to what a child says, and in smart digital assistants on smartphones (such as Siri) or smart speakers (such as Google Home or Amazon's Alexa). These AI-based applications present numerous opportunities, and also risks, for children regarding their well-being and privacy. How do children experience artificial intelligence? Are they aware of the possible risks and to what extent do they understand how AI works? 

The insights in this bitefile are based on research with children between 4 and 16 years old. The exact age group differs per study. The bitefile was made in collaboration with the Dutch Media Literacy Network. All consulted literature can be found here.

The risks of artificial intelligence for children

When talking about the risks of AI, many people think of scenarios where robots take over the world. Fortunately, that is not (yet) reality. There are, however, several other more realistic risks regarding AI that children face.

• AI poses risks to a child’s privacy. Many AI driven applications store personal data that is then used by companies to improve their services, advertisements, and products. Children are often not aware of that data-storage or the fact that advertisements/services are finetuned to their shown interests and needs. Research shows, for instance, that children while playing with smart dolls are not aware of the fact that all their conversations are stored and can be used by the doll’s creators for other purposes. Children also don’t realise that the reply of a toy is not a real response to their question or comment, but a pre-programmed script. 
• AI can cause social inequality. AI technology is often expensive, and not every family has the financial means to introduce their children to AI at an early age. In addition, having and not having the latest gadgets can create a social hierarchy among children, increasing the risk of bullying.
• Using social media platforms with AI-driven smart algorithms (such as Netflix, YouTube and Facebook) can be addictive because these algorithms aim to keep users on the media-platform as long as possible. These algorithms succeed in doing this by constantly offering new, personally relevant series, videos, or information.
• Another risk of such AI-driven personalized media offering lies in the filter bubble. Children see and hear more of the same thing and are, thus, exposed to one-sided information.
• AI can be used to make deepfakes, also known as fake videos. These videos are almost indistinguishable from the real thing and are often full of misinformation. Children can be misled by these types of videos.

The opportunities of artificial intelligence for children

There are several ways AI can enrich children's lives and course of development:

• AI offers new possibilities for education, provided that the privacy of the user is handled properly. For example, AI can be used in the classroom to closely follow the learning process of students by monitoring digital assignments and tasks. Researchers also see great potential in the use of AI-driven digital tutors that can support students in their personal needs, enabling personalized education and guidance.
• According to the recent developments in the field of youth healthcare, AI techniques create new opportunities for wellbeing and mental health. For example, smart computer games and robots can be used to better support children with disorders and behavioral problems. Research shows that social robots can help children with social disorders to develop social-emotional skills. By interacting with these robots, children build a bond with them. This bond makes it possible to practice and train their social behavior, such as empathy, recognizing emotions or even imitating social situations, in a safe (social) environment.
• Children can even develop a true friendship relationship with robots and other social AI applications. A robot can feel like a real friend who cares and whom they can talk to. To develop a friendship with a robot, it is important that a child feels connected with the robot, trusts it, and feels the robot can support them. 
• Interaction with AI can prepare children for the future. AI can no longer be ignored when thinking about the future society, and will, probably, to a large extent how we live, learn, work, and communicate in the future. By educating children about AI technology, raising awareness of the opportunities and risks, and introducing them to programming, we prepare a new generation to use and develop technology responsibly.
• Interaction with AI can increase children's interest in the subjects that underlie AI, such as engineering, mathematics, computer science, and psychology. With the right guidance, they can become enthusiastic and maybe even passionate about these fields and about pursuing an education and professional career in one of them. 

How do children experience artificial intelligence and what do they understand about it?

Research shows that children often do not realize that their favorite apps and games use AI technology behind the scenes (or rather, screens). They are therefore not always aware of their interactions with AI. Research further shows that the following four elements are central to the way children experience AI technology:

Perceived intelligence

• When children interact with AI-driven products, such as smart toys, social robots, and smart speakers, they make an estimation of the intelligence of the product. This “smartness” of a product is judged based on the "behavior" of a product. This assessment contributes to the realization of what the product can and cannot do. For example, children experience self-driving cars as relatively stupid, but they consider a Google speaker to be just as smart or even smarter than themselves.

Identity

• Children assign a certain identity to AI-controlled products. They can interpret AI-driven behavior as if it is a real personality with human characteristics. This mainly happens with smart toys, social robots, and the main characters in interactive video games.

Playfulness

• The playfulness of an AI-driven product is important for children. Most children see AI-driven products as a partner in crime that always wants to play with them, whereas adults and peers do not always want to. The more playful a product is, the more value the child will attach to it and the deeper the child bonds with the product.  

Understanding

• Children who better understand how AI works, make a better assessment of what AI-driven products and apps can and cannot do. They also assign greater intelligence and a stronger identity to AI products and apps. Younger children tend to understand AI technology less than older children. However, when technology is explained to young children in an accessible way, their knowledge grows rapidly, and they can make a more realistic assessment of what AI can and cannot do. They can better recognize AI and they are more aware of the risks that AI carries.

A glimpse into the future

AI is a "hot topic" in scientific research. Yet many important questions about children and AI remain unanswered:

• Relatively little is known about the effects of AI and interaction with AI-related technology on the health and well-being of children. Big question marks remain in different fields, especially within neurophysiological and social research, such as: what effect does AI have on the development of the child's brain? And what influence does AI have on the human social skills of children?
• There are also ethical issues for which there is no clear answer yet. Is it ethical to outsource educational and educational tasks to AI-driven machines? Can data from children be used to improve products and services, even if it involves something positive such as helping others better?

What we can do to help children pick the fruits of what artificial intelligence has to offer

Research shows that parents and teachers play an important role in the way children experience AI and the extent to which they understand it. They can help children take advantage of AI’s opportunities and minimize their risks. How do you do that?

• First, try to understand AI yourself and have an idea of ​​the opportunities and possible risks. This can be done by looking up information about the topic, there are many information websites and free courses online.
• Talk to children about the opportunities and risks of AI. Explain to them what AI is and how it works. If possible, supervise children during their first interactions with AI and give feedback. 
• Remember: every child is unique. How children deal with AI and the extent to which they grasp the subject is different for each child. It is important not to generalize the risks and to consider each child’s individual situation.
• Have children watch children's shows and movies about robots (such as Disney-Pixar's Wall-E) or have them read stories about AI. These stories and programs are tailored to a child's way of thinking and make topics such as robotics and AI understandable. By doing so, they can help to start a conversation about AI. Keep in mind that fictional stories about AI, such as robots, can cause children to develop an unrealistic view of what AI is and what it can do. Therefore, ask children for their opinion and ideas about AI and adjust their image about AI if necessary.