Parenting in the Age of AI

More young people are using AI than ever. Early research indicates some reasons to be concerned, but how do we protect our kids without letting them get left behind?

If you have a teenager at home, the odds are that they are using generative AI. Recent research suggests that 7 out of 10 US teens have used AI companion tools, and over half of US teens talk to an AI companion multiple times a month. If that comes as a surprise to you, you are not alone. In households with kids using AI, only 37 percent of parents actually know that it’s happening in the first place.

Many parents are worried about the impacts of AI on their children’s safety, with over 75 percent of parents expressing concern about the threat AI might pose to their children. But before we start taking away any devices, it’s important to first understand the reasons and risks behind child AI usage, and the practical things parents can actually do to keep their kids safe and flourishing.

What are kids using AI for?

While it’s a common belief that kids primarily use AI to cheat on their homework, the truth is that young people are using AI for a variety of purposes including getting information, brainstorming, making images and music, and even coding.

Kids are of course using AI as a tool to help with schoolwork, but they are also seeing a wide range of benefits from AI across many aspects of their lives. And as they grow up in a world that demands increasing technical ability, getting to use these tools at a young age could be an advantage when it later comes time to work in jobs that encourage active AI use.

What are the risks of using AI?

Despite the benefits that AI can offer, there are various concerns that parents should be aware of for their children’s sake.

Misinformation and Misuse

By design, generative AI models are adept at replicating human-made text, speech, images, and videos. In the hands of bad actors, these capabilities can be used for manipulation and deception, such as impersonating individuals through realistic deepfake videos, facilitating grooming efforts, or enabling online scams. With these capabilities, generative AI models are increasingly being used to create non-consensual intimate imagery (NCII), and research shows that children are a growing target for deepfake pornography.

In addition to fabricated sexual imagery, misleading and false information created by generative AI is prevalent and growing across the internet. Although these tools can also be used as a means to detect and prevent misinformation, the ease of creating personalized content opens up a large-scale potential for fraud and manipulation.

Emotional Impacts

Another ongoing concern around AI is the impact of social companionship and dependence. While some use of AI companions has shown to be beneficial and provide social support, use of AI companions can displace relationships with real friends and family. At an extreme, the social interaction between kids and AI can even devolve into an emerging concern known as AI psychosis, where users enter a psychotic state associated with intensive and prolonged use of AI.

Extreme psychological consequences appear to be rare occurrences, and it is not clear that AI use is itself causing psychological harm to its users. Some research has even suggested that mental health problems cause AI dependence rather than AI being the cause itself. However, the consequences of AI psychosis can be severe: multiple teenage suicides have been reported in relation to problematic AI chatbot usage.

Despite the amount of attention AI psychosis has garnered in popular media, research on the emotional impact of AI use is still in its very early stages. Because the ability to form relationships is critical to adolescent development, it’s worth devoting continued attention to the long-term consequences of AI use by children.

Cognitive Offloading

As kids (and adults) increasingly use AI across their personal, school, and work lives, AI is doing more and our brains are doing less. Just as using GPS can make us forget how to read physical maps, there are concerns that AI is causing a phenomenon known as “cognitive offloading”: the effect of outsourcing demanding mental tasks, which could impact skills such as creative ideation, writing, organization, and planning.

Although the long term psychological consequences of AI use are currently unknown, early research suggests that cognitive offloading to AI may inhibit certain essential skills such as critical thinking. Researchers at MIT similarly found that relying on AI tools actually reduced brain activity in essay writers compared to students not relying on AI. Of course, the use of AI for personalized learning can be a great benefit for education. However, the possibility of harm from cognitive offloading means the implementation of any AI tools in education warrants careful supervision.

What can we do?

You Can’t Ban Usage in a Vacuum

While the thought of AI leading to grooming, psychosis, and a loss in critical thinking might seem to point to an outright ban on AI use for kids, the data suggest otherwise. Studies on reducing child risk from social media found that engaging in active parental guidance better prevents risk than strict usage restrictions. For example, rather than simply taking away access to devices or platforms, parents are likely to find greater success through proactively explaining potential risks and sitting side-by-side with their child, showing them how to navigate the platform safely and securely.

While some AI platforms plan to give parents greater control over their child’s AI use, a part of the challenge in implementing strict usage bans on kids is that the widespread nature of AI technology means that actually enforcing those bans can be very difficult.

Support Community Controls

Although parents may find it difficult when acting alone, strict age restrictions are something that can be enforced by lawmakers and have already begun to be implemented by some AI companies. Without legal backing, individual parents likely face an uphill battle, but with strict age verification policies, AI companies themselves would be forced to monitor the gateways of AI use and provide a backstop to prevent child access. And, even before the pen is put to paper by lawmakers, pressure by parents and community leaders can take effect, causing AI companies to preemptively implement policies in anticipation of future legislation. If you are interested in widespread usage restrictions for children, supporting legislation in your local, state, and federal government could lead toward a long-term solution.

Another way to secure child safety is to directly consult other parents and teachers in your community. While AI is also used in the home, AI use in schools is rapidly growing. Asking teachers and other parents about the policies they have in place in the classroom and the home can help foster consistent guidance that spans from the household to the classroom. Parents, when acting together, have already found success in these community-based efforts such as the creation of phone-free school policies.

Start the Conversation Early

Despite many of the concerns parents have for their children's AI use, 43% of parents haven't talked with their kids about how to recognize AI-enabled threats. Although these conversations might be difficult and imperfect, educating children early about the risks of AI can help children understand and safely employ the technology.