Analysis: Trust in the technology for young people appears to be borrowed in many cases from trust in existing human relationships

Artificial intelligence is rapidly making its way into schools, healthcare and social services. The Irish National AI Strategy is heavily invested in bringing AI tools into public services and education, promising greater efficiency, accessibility, and support. Yet amid the excitement around technological innovation and commitments to “human-centred AI”, an important question is often overlooked: what makes AI purposeful and trustworthy in the first place?

Much of the current debate starts from two assumptions about young people. The first is that they are “digital natives” who naturally embrace services delivered by new technologies because they have grown up surrounded by them. The second is that they are particularly vulnerable to technological harms, misinformation, and manipulation. Both assumptions shape discussions about AI in education and support services.

However, my recent research with young people suggests a more nuanced reality. Over the past year, I have worked with young people aged 15 to 18 attending the Glen Youthreach programme in Cork. It provides education, training, and support for early school-leavers who have often experienced educational exclusion and other forms of social or psychosocial disadvantage. Through workshops and conversations, we explored young people’s perceptions of chatbots and generative AI, including where they might be useful and where they should not be used.

We need your consent to load this rte-player contentWe use rte-player to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences

From RTÉ Radio 1’s Brendan O’Connor, can a chatbot be a good friend? Two Gen Z guinea pigs find out

While participants recognised potential benefits, particularly around convenience, accessibility, and support outside normal service hours, few expressed unconditional enthusiasm. Many were sceptical. Some described chatbots as “fake”. Others questioned whether AI could genuinely understand human experiences. Several participants explicitly rejected the idea of discussing personal or emotional problems with a chatbot. Yet this scepticism did not amount to a rejection of technology.

The young people demonstrated thoughtful views about where AI might have a role and where it should not. They could identify situations where chatbots might be useful for information, signposting, or low-stakes support. Equally, they articulated clear limits. Emotional support, trust, and difficult life experiences were repeatedly described as matters that belonged within human relationships.

At first glance, these findings might appear to be about trust in technology. However, a closer look revealed something more interesting. The young people rarely discussed trust as a property of the chatbot itself. Instead, trust was consistently connected to people, places, and relationships. When participants imagined using a chatbot in their educational setting, their willingness to engage depended heavily on their trust in the staff members who introduced it. A chatbot endorsed by a trusted youth worker or educator was viewed differently from the same chatbot introduced by an unfamiliar organisation. Trust in the technology appeared, in many cases, to be borrowed from trust in existing human relationships.

A chatbot endorsed by a trusted youth worker or educator was viewed differently from the same chatbot introduced by an unfamiliar organisation.

Even where participants were open to using AI tools, they consistently preferred support from trusted professionals. Where trust in people was absent, the technology itself had little chance of gaining legitimacy. Equally important was the role of place and belonging. For many participants, Youthreach represented far more than an educational programme. It was a community, a safe environment, and a place where they felt accepted after often difficult experiences in mainstream education. Discussions about AI were therefore never only discussions about technology.

They were also discussions about what kind of place their programme should remain. Some participants worried that introducing chatbots could signal a reduction in human support. Others were concerned that technology might undermine the relationships, informal conversations, and trusted adults that they valued most. In these cases, resistance to AI was less about technical risk than about protecting existing social relationships.

These findings suggest that trust is not simply an attitude that can be engineered through better design or more sophisticated AI models. Nor is trust likely to emerge through top-down implementation strategies that treat AI adoption as inevitable. Rather, trust emerges through what might be described as relational infrastructures: the networks of relationships, practices, institutions, and shared expectations that make support systems function.

We need your consent to load this rte-player contentWe use rte-player to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences

From RTÉ Brainstorm, Dr ChatGPT will see you now: how AI could be bad for your health

From this perspective, AI implementation is not primarily a technical challenge. It is a social and organisational one. This distinction matters because current policy discussions often focus on making AI safer, more transparent, more explainable, or more accountable. These are important goals. However, they can create the impression that trustworthiness is primarily a property of the technology itself.

My research suggests that what makes an AI system seem trustworthy and purposeful may depend just as much on factors that sit outside the technology. Who introduces the system? Why is it being implemented? What problem is it trying to solve? How does it affect existing relationships? Do young people have genuine influence over decisions about its use? Is the technology perceived as complementing human support or replacing it?

These questions may matter more than the sophistication of the underlying AI model. They also encourage a broader conversation about whether AI is necessary in the first place. In some contexts, AI may help extend access to information or support. In others, investing in additional staff, stronger relationships, or better services may be the more appropriate response. The answer should not be predetermined by technological enthusiasm or assumptions about inevitability.

Before asking how we can make young people trust AI, we may need to ask a more fundamental question: what purpose should AI serve in their lives at all?

The findings also highlight the importance of involving young people in decisions about technology before implementation occurs. Participation has long and frequently been discussed in relation to designing AI systems. Much less attention is paid to whether young people should have a say in whether such systems are introduced at all. Yet the young people in this study demonstrated valuable expertise about the environments in which these technologies may operate.

They identified potential benefits, recognised limitations, and articulated concerns about the preservation of human relationships and community. In doing so, they acted not simply as future users of AI systems but as stakeholders with important knowledge about what meaningful support looks like.

As governments, schools, and social services continue to develop AI strategies, the debate should not begin with what AI can do. It should begin with what problem needs solving, whether AI is the most appropriate response, and how those most affected understand the trade-offs involved. Before asking how we can make young people trust AI, we may need to ask a more fundamental question: what purpose should AI serve in their lives at all?

Follow RTÉ Brainstorm on WhatsApp and Instagram for more stories and updates

The views expressed here are those of the author and do not represent or reflect the views of RTÉ