The future of Artificial Intelligence

In today’s rapidly evolving technological landscape, Artificial Intelligence (AI) has the potential to reshape industries, systems, and even societies. But what does this mean for children? As AI becomes increasingly embedded in our lives, its implications for the rights, well-being, and futures of children need to be addressed. We had an intergenerational conversation with two inspiring leaders, Selin Ozgursoy, a UNICEF Youth Foresight Fellow who is 18 years old, and Bo Viktor Nylund, Director of UNICEF Innocenti, to explore the opportunities AI presents for children and young people, and what the future may hold.

Interviewer: Artificial Intelligence has been around for a while now, and discussing the risks and opportunities, especially for children is long overdue. Thank you for being here, Bo Viktor and Selin. Let’s kick off the conversation with you Selin, How do you interact with AI daily, what are your go-to tools?

Selin Ozgursoy: I use AI in many aspects of my daily life, starting with tools like ChatGPT, which I interact with frequently. I also engage with AI through voice assistants like Siri and Alexa, Netflix recommendations, and security systems. Beyond everyday use, I’m interested in researching AI models, especially large language models and machine learning. I am particularly passionate about exploring AI’s potential in fields such as sustainable transport, cybersecurity, and quantum computing, and aim to work on helping shape a better future with AI.

Interviewer: Bo Viktor, what opportunities do you think AI provides for children, especially in areas like education, safety, and health?

Bo Viktor Nylund: Two key opportunities immediately come to mind. First, in healthcare, AI could act as a digital 'twin' for parents, helping them keep track of their child's health. It could remind them about things like vaccinations, balanced diets, or health milestones. For pregnant women, it could offer tips and reminders as they prepare for their baby’s arrival. AI might even assist with minor issues like treating a bruise, although it’s always essential to seek professional medical advice when needed.

Second, in education, AI holds even greater potential. In our recent publication on the Future of Childhood in 2050, we highlight a growing education gap, especially in Africa, where population growth is outpacing the education system’s ability to keep up. AI-powered tutoring could be revolutionary, offering personalized learning experiences to bridge this gap. While it would need to be fully integrated into broader education systems, its potential to improve student outcomes is massive.

Interviewer: Selin, do you have a different perspective on the opportunities AI presents for children?

Selin Ozgursoy: AI can really make parental care easier, helping them track their children’s learning progress or health, as Bo Viktor mentioned. Education, for instance, will become more personalized and interactive, with AI offering customized learning tools for children with specific needs. Additionally, AI could help set a universal learning standard—not that every child everywhere will learn the exact same things, but a foundational understanding of universal rights, ethics, and societal norms. This could contribute to closing educational gaps globally. Then there’s safety: AI can enhance both online safety, like filtering harmful language to prevent cyberbullying, and physical safety: providing real-time alerts during natural disasters, and protecting children from harm.

Interviewer: Selin, have you come across specific examples of biases in AI, and how do you think these biases affect children?

Selin Ozgursoy: Yes I have. The biases in AI stem from its creators, because, as humans, we all carry biases—whether personal or cognitive and that’s a problem. If AI continues to be developed by a small, homogenous group, its decisions will reflect their limited experiences. We need to ensure that AI is trained on diverse datasets that represent children from different backgrounds. Transparency and inclusivity are key. AI in public services can struggle to address the specific needs of minority groups, resulting in unequal access to vital resources. This, combined with the digital divide, where marginalized children face barriers to AI technology, can exacerbate social and educational disparities. However, I believe these biases can be addressed.

Interviewer: Bo Viktor, kindly share your thoughts on the biases Selin mentioned, particularly about how they affect children's rights. What do you think can be done to mitigate these biases?

Bo Viktor Nylund: That’s a great question! Before even addressing bias, we must consider access to the internet. Many children, especially in the global South or rural areas, face barriers that prevent equal access. Without access, discussions about bias become secondary. When we discuss bias, one key area is language. It's crucial to teach children and young people how to recognize biases in the systems they use, as these biases already exist and will persist. 

To mitigate these biases, Governments, organizations, and the private sector must work together to ensure AI systems respect children’s rights, privacy, and security. There’s an ongoing debate about solutions like watermarks to distinguish AI-generated content, but one thing is clear—we must contribute to making these tools genuinely safe for kids. For AI to truly serve children, we need robust safeguards.

Interviewer: Thank you, Bo Viktor. Would you like to contribute your thoughts on this, Selin?

Selin Ozgursoy: Yes! Bo Viktor highlighted key points about ensuring diverse and inclusive data sets for AI. We must prioritize digital literacy across all sectors—parents, educators, policymakers, and the private sector. And while the private sector leads in AI development, true progress will come from collaboration between public and private institutions, ensuring equity and sustainability. We also need to engage children and youth in AI development so they can help design the tools that protect their rights.

Interviewer: Thank you Selin for those insights! The conversation about AI's impact on children is subtle. Bo Viktor, how can we ensure that children's rights are protected concerning AI, particularly regarding accountability, transparency, and safety? 

Bo Viktor Nylund: First, listening to children and including their views in our research on online engagement, safety, and protection is crucial. UNICEF’s 2022 AI guidance provides ethical recommendations for governments and the private sector to adopt child-centered approaches, especially with generative AI. Second, while governments play a significant role in enforcing justice, the private sector must also ensure responsible AI development. Instead of creating a new body, I believe we should leverage UNICEF’s existing expertise to keep children’s voices at the forefront and hold the organization accountable for advancing this critical agenda.

Interviewer: Do you have concerns about the future of AI for children and their rights, especially looking ahead to 2050? 

Selin Ozgursoy: I believe children and youth should play a pioneering role in AI development. To reduce AI risks, we must ensure human oversight—humans and AI together are more powerful than either alone. Education is key; we should integrate AI into school curricula, ensuring it is age-appropriate. Starting with basic programming in elementary school and moving to machine learning and ethics in high school would be beneficial. Keeping humans in the loop and providing opportunities for young voices is essential for a sustainable future.

Bo Viktor Nylund: I see both opportunities and risks with AI. On the positive side, generative AI could help reduce global inequalities if used correctly, and UNICEF has a critical role to play in this. However, one major concern is AI gaining too much independence, and making harmful decisions. It’s essential to set clear limits on what AI can and cannot do, as this poses real risks we must address proactively.

Interviewer: What is your vision for the future of AI for children in 2050?

Selin Ozgursoy: By 2050, I envision a world where AI has addressed its imperfections and biases, becoming a valuable tool for children in education, healthcare, and online safety. I hope we see no power dynamics or conflicts over AI leadership, fostering a strong partnership between the public and private sectors to ensure inclusive and diverse products. My vision is optimistic, and believe it can only be realized through the efforts of young visionaries committed to using AI for good. 

Bo Viktor Nylund: By 2050, I believe generative AI will allow governments to tailor services to children’s unique needs, empowering families to make better decisions using AI tools. This will help children themselves address their own developmental needs.