SHIFT-AI: Bringing Young People’s Perspectives into the Future of AI
AI is already part of everyday life for many young people. It shows up in how they study, how they communicate, and how they make decisions. But while these systems are moving quickly, the question of how safe they actually are for younger users is still catching up.
SHIFT-AI (Safeguarding Human-AI Interaction via Feedback from Teens) is built around that gap. The project brings young people directly into conversations about generative AI, not as participants being studied, but as people actively shaping what safer systems should look like.
Led by the Design Bioethics Lab at the University of Oxford’s Department of Psychiatry, and supported by the UK Government’s AI Security Institute (AISI) Challenge Fund, the work focuses on understanding how AI is really being used, and what that means for risk, safety, and responsibility.
In-Person Session
Alongside several online discussions, the project also included an in-person session with members of the NeurOX Young People’s Advisory Group (YPAG). It gave people the chance to come together, reflect on the earlier sessions, and talk more openly about their own experiences in a setting that felt less formal.
The focus was on listening to what young people are already doing with generative AI. They spoke about using it for schoolwork, social media, and everyday tasks, and shared what they find helpful as well as what does not sit quite right.
What stood out was how mixed these experiences can be. The same tool might help with ideas or productivity, but also raise doubts about whether the information can be trusted, how accurate it is, or how much influence it has.
During the session, the group then worked on turning these reflections into something they could share with their peers. They began developing a toolkit comprising social media posts, short-form videos, and short quizzes.
These were built around situations they recognised from their own lives, using simple prompts and questions to encourage people to pause and think about how they are using AI. The focus was on making something that feels relatable and easy to engage with, rather than technical or distant.
Image: Participants and facilitators at the SHIFT-AI in-person session
Hearing from a YPAG Member
“For me, being part of the SHIFT-AI project has been a unique experience that has allowed me to feel like my perspectives as a young person have been not only taken into consideration, but in fact used to shape discussions and research in the AI world. It’s very common for young people to have our viewpoints considered only in a tokenistic way, but here I have felt a genuine respect for our insights.
I’m someone who’s very skeptical of AI, but in our team there have been some who are more welcoming to it, as well as those who are even more strongly against its use than I am. This has been very enlightening and, I think, has allowed us to consolidate an intellectually diverse and eclectic collection of perspectives. It has also just been fun, particularly at the in-person event where we were able to use all of the resources we had created to plan and create the Toolkit. We created questions for quizzes, designed infographics and created and voice acted for a safety video.
I think that young people are the ones whose lives are going to be the most impacted by AI, so having our perspectives inform our risk model, mitigation models and toolkit allows these resources to reflect our concerns in a way that a lot of approaches toward AI lack. Some things that we’ve talked about haven’t been as much of a focus in a lot of mainstream discussions about AI so I think it’s very important that these issues are considered, which I hope the project enables.
In terms of impact, I do think that we have a lot of opportunity for what we’ve made to take off on social media. The video and quizzes could all be used in school assemblies so I think, since our content was made by us and therefore represents the concerns of young people, there’s a strong chance that it will resonate with many others our age. This could have a large impact on the level of caution with which other young people approach and use these technologies.” – Finn McGrath
Why this approach matters
A lot of work on AI safety still centres adult users or technical perspectives. While that is important, it does not always reflect how younger people are actually engaging with these systems.
Their use of AI is shaped by where they are in life. It connects to how they learn, how they build relationships, and how they make sense of information. These factors can change both the kinds of risks they face and how those risks play out.
By working with young people as part of the research process, SHIFT-AI is able to surface concerns that might otherwise be missed. It also makes it possible to think about safeguards that feel realistic, rather than purely theoretical.
Looking Ahead
SHIFT-AI is still ongoing, and its direction continues to be shaped by the young people involved.
The aim is to understand the risks properly and use that insight in a way that actually leads to change, whether in how these systems are designed, how guidance is written, or how policy is shaped.