Glossary of terms

Before delving into the findings of this commentary, we first want to clarify terminology.

Digital technologies: Digital technologies are expansive and can mean different things to different people. In this commentary, we consider the term in the broadest sense possible, capturing:

• Hardware such as the physical devices people use, including smartphones, tablets, smartwatches, e-readers, laptops and desktop computers.
• Software – both system software needed for hardware to function (for example, operating systems on smartphones) and application software that help you carry out specific tasks on a piece of hardware (for example, email software, online applications that we can download ourselves).
• The data needed for many digital technologies to work (for example, personal data captured by online platforms, or operational data monitoring software performance).

Young people: As much as possible, we reference the specific age range each piece of evidence relates to. When speaking in general terms, ‘young people’ refers to individuals between the ages of 14 and 24.

Mental health and well-being: In this commentary mental health refers to emotional, psychological and social well-being that affects how individuals think, feel and act.

Identity: Identity, similar to digital technologies, is an expansive term. In this data commentary, we consider identity in terms of how young people come to learn about themselves and their value, as well as how they choose to present themselves to others.

Social media: We refer to social media as an umbrella term for a range of internet-based platforms that predominantly feature user-generated content (including images, videos, discussions, messages). Common platforms include Meta’s Instagram and Facebook, X (formerly Twitter), TikTok, YouTube, Snapchat and Reddit. These platforms usually have some distinct functionalities from one another and tend to be engaged with in different ways. In this commentary, we use this umbrella term where available data does not delve into the specificity of different platforms but recognise that as a general term it is limited in capturing the heterogeneity of online experiences.

Large language models (LLMs): LLMs are a type of artificial intelligence (AI) system trained on text data that can generate natural language responses to inputs or prompts[ref]Analysis and Research Team (2023). ChatGPT in the Public Sector – Overhyped or Overlooked? Council of the European Union General Secretariat. Available from: https://www.consilium.europa.eu/media/63818/art-paper-chatgpt-in-the-public-sector-overhyped-or-overlooked-24-april-2023_ext.pdf. [/ref]. A popular example of this technology is ChatGPT, a tool that can be used to summarise notes, generate written content and provide answers to questions, among other things[ref]Jones, E. (2023). Explainer: What Is a Foundation Model? Ada Lovelace Institute. Available from: https://www.adalovelaceinstitute.org/resource/foundation-models-explainer/. [/ref].