Children need protecting from social media and generative AI | Letters

"We in the Neuroscience, Ethics and Society (Neurosec) team at the University of Oxford contend that such investigations should extend also to whether young people should have access to generative AI. In the case of social media, ministers and commentators have focused on features like addictive feeds and age limits; our research team's work with young people shows that we must also reckon with such considerations among many others in an era of AI-driven technologies."

"AI-based chatbots are increasingly present in young people's lives across a host of domains, from education to companionship. And adolescence is a formative stage for developing social understanding, one's sense of identity and so on. This raises urgent questions, such as: at what age should young people have access to AIs simulating friendship or intimacy?"

"Any policy focused on protecting children must broaden its scope beyond traditional social media platforms, including all relevant technologies that affect young people's development. Consultations should invite evidence from developmental psychology and AI ethics, and perspectives from young people themselves. Asking simply whether a platform should be offlimits misses a deeper question: what kinds of relationships to technology do we want future generations to form, and at what age is it appropriate to begin them?"

Investigations into banning social media for under-16s should also examine young people's access to generative AI. AI-based chatbots appear in education, companionship, and other domains and can shape social understanding and identity during adolescence. Concerns about mental health, social comparison and addictive design apply but are insufficient alone. Additional risks include manipulation, dependency, and artificial intimacy created by AI-simulated relationships. Policy aimed at protecting children must broaden beyond traditional platforms to include all technologies affecting development. Consultations should incorporate evidence from developmental psychology and AI ethics and include perspectives from young people. Policymakers must decide the appropriate ages for different technological relationships.

Read at www.theguardian.com