How do teenagers use AI – and what shapes their trust?

The project is led by Dr Ilaria Torre at Chalmers and Dr Marta Romeo at Heriot-Watt University. Together, they aim to contribute new knowledge about young people’s trust in technological development.

Are you an upper secondary school teacher in Gothenburg? Sign up via the link: Focus groups on students’ use of AI

Focus on trust, overtrust, and distrust

From chatbots such as ChatGPT which can help with homework, to AI assistants that offer advice or answer questions - these technologies are becoming a routine part of young people’s lives. Yet despite their growing presence, we still know surprisingly little about how much teenagers trust these systems, when they rely on them too heavily, or when they avoid them altogether.

The researchers aims to uncover how trust, overtrust, and distrust in AI can develop, and how these behaviours differ depending on socio economic background.

The project is an international collaboration with teams located in Sweden and Scotland, and teenagers from different socio-economic backgrounds will be recruited, allowing researchers to gain insights into whether young people from different circumstances use and trust AI in different ways.

“We are really excited to kick-start this international project, the AI-hype is at its peak and we believe it’s vital to understand how people are actually experiencing it” says Dr. Torre.

Building better AI for young people

The findings will support schools, parents, policymakers, and technology developers in creating safer, fairer and more age appropriate AI systems.

The project will produce practical guidelines to inform the responsible design and use of AI tools for teenagers from all backgrounds, which will be shared with local, national, and international authorities, such as local high schools, skolverket, the Scottish AI Alliance, and AI Sweden.

Timeline and implementation

• Technical development (January - May 2026). The researchers have been developing a secure “digital diary” application, that lets participants use ChatGPT as they normally would, but allowing researchers to access their conversations with the AI. Through the application, participants will also be invited to reflect on what they used the AI for and how helpful or trustworthy they found the responses.
  
  The application safely stores data in an anonymised way, so that only researchers can access it, following GDPR regulations. Participants will be fully briefed on the purpose of the app, and will give informed consent before data collection begins. Experts in AI ethics have been consulted to ensure that any potential risk is minimised, and that appropriate protocols are in place for correct data handling.
  
  
• Survey to parents and teachers (March 2026). The researchers sent out an initial survey to parents and teachers of teenagers, in both Gothenburg and Edinburgh, to understand their perceptions of how teenagers are currently using AI.
  
  
• Focus groups with teachers (April-May 2026). The researchers are planning to host focus group meetings with secondary school teachers (in both Gothenburg and Edinburgh), to share with them results from the survey, and also to get feedback on the project and on the digital diary prototype. 
  
  Are you an upper secondary school teacher in Gothenburg? Sign up via the link: Focus groups on students’ use of AI
  
  
• Data collection with teenagers (September 2026). Secondary school students (aged 16-18) will be recruited from Gothenburg and Edinburgh, and will be invited to use the “digital diary” app over several weeks.
  
  
• Guidelines development (October-November 2026). The researchers will analyse the teenagers’ conversations with ChatGPT, specifically focusing on any differences based on social background. Results will form the basis of suggestions for policy guidelines that will be shared with relevant stakeholders, with the aim of updating and improving policies for teenagers’ use of AI.

The team

The team is composed of an interdisciplinary mix of specialists, who have backgrounds across Human-Robot Interaction, AI, Virtual Companions, Trust, Computer Science, Linguistics, Youth Engagement, Design and Engineering.

•    Dr. Ilaria Torre, research lead (Gothenburg team)
•    Dr. Marta Romeo, research lead (Edinburgh team)
•    Dr. Carl Bettosi, research technician
•    Negin Hashmati, PhD student
•    Camilla Irvine-Fortescue, PhD student
•    Nelly Hallgren, research assistant
•    Razvan Albu, research assistant

The project, officially called “D-AI in the life: Understanding AI Use Among Teenagers in Unequal Contexts” has received funding from a Google Academic Research Award for one year.