Millions of young people and educators around the world use our technology to teach, learn, and create with code. Building it safely and responsibly is intrinsic to who we are and what we stand for.

A growing body of research and public commentary links the use of social media, smartphones, and algorithmic platforms to declining mental health and attention in young people. Public attitudes are shifting in response, and parents, teachers, and governments are increasingly demanding action. Phone bans in schools and restrictions on social media are spreading across many countries. The technology sector can no longer take public trust for granted.

None of this is new territory for researchers, who have argued for decades that technology is not neutral, and that the choices made in designing it can either embed harm or help prevent it (1). We think this moment calls for a calm, evidence-led response, and we want to be open about how we are approaching it and to do this work alongside others rather than on our own.

Standing on existing research

We have always been a research-led organisation. When we face a difficult question, we look at the evidence and learn from the people who have studied it.

Children’s digital rights is a rich and well-established field. Researchers and experts have spent years thinking carefully about what children need from the technology they use, and how their rights apply in a digital world. The United Nations set out a clear framework in 2021 with General Comment No. 25, which describes how children’s rights apply to the digital environment. Organisations like UNICEF and the 5Rights Foundation have built on this with practical guidance for the people who design and build products.

This body of work gives us firm ground to stand on. Rather than starting from scratch, we have used it to shape our own approach. We have drawn in particular on UNICEF’s Responsible Innovation in Technology for Children framework, UNICEF’s EdTech for Good Framework 1.0, the UN’s General Comment No. 25, and the Digital Futures Commission’s Child Rights by Design principles.

Eight principles to guide our work on edtech

We have begun by writing eight principles, designed to guide the decisions we make as we build and run our products.

Our principles:

1. The best interests of the child come first. We design for children’s wellbeing, which means more than safety and privacy. It includes their agency, their emotional health, their relationships, and the space to create. Sometimes putting that first means choosing against growth, engagement, or speed.

2. Our technology supports human relationships. It does not replace them. Any personalised or AI-supported features we build are there to strengthen the relationships between young people, educators, and caregivers. People stay in control, and we do not hand decisions about a child’s learning or wellbeing to a machine.

3. We only collect the data we need. By the time a child turns 13, more than 72 million pieces of personal data will have been collected about them. We collect the minimum we need to help children learn and to measure our impact. We do not sell data, and we never will.

4. We design for learning, and we prove its impact. Our products are grounded in evidence about how children learn, and we evaluate whether they actually work. We avoid designs that keep learners hooked rather than helping them learn.

5. We design with children and educators, not for them. The people who use our products are the experts on their own needs. We involve them in research, testing, and feedback, and we make sure what they tell us shapes the product, not just a report at the end.

6. We design for diverse needs, abilities, and contexts. Our technology adapts to people, not the other way around. We do not assume constant connectivity, confidence with digital technologies, or a classroom that looks like anyone else’s.

7. We are transparent and accountable. We use plain language to explain how our products work, how data is used, and what rights people have. And we give people clear ways to question or challenge the decisions our systems make.

8. We apply the highest standards of children’s rights everywhere we work. Where local rules fall short of international best practice, we choose the higher standard. A child’s rights should not depend on where they happen to live.

You can read the full set of draft principles, and the research behind each one, in this PDF:

Committed, not perfect

Publishing these principles is a statement of intent, not a claim that we have got everything right. This is a fast-moving area. Regulation is changing quickly, the research keeps developing, and we grow and change our own products. There are areas where we know we have work to do, and almost certainly areas we have not yet spotted.

These principles are how we hold ourselves to account, and they are only a start. Alongside them, we are mapping the regulatory landscape across the countries where we work, developing ways to assess our products against these principles, and building these commitments into how we make decisions. We will keep listening to the people who use our products, and making sure children themselves have a voice in shaping the technology built for them.

We also know we cannot do this alone, so we have two asks for you:

If you have a view on these principles, tell us. Whether you are a researcher, educator, parent, young person, or policymaker, we want to hear where you think we have got it right, where we have not, and what we have missed. Honest challenge is genuinely useful to us. Give your feedback and sign up to be involved in shaping them.

If you are part of a non-profit or organisation working on similar questions, get in touch. We would like to build a coalition of like-minded organisations to share what we are learning and raise standards across the sector.

---

(1) See for example:

• Friedman, B., & Nissenbaum, H. (1996). Bias in computer systems. ACM Transactions on Information Systems, 14(3), 330–347. https://doi.org/10.1145/230538.230561
• Noble, S. U. (2018). Algorithms of Oppression: How Search Engines Reinforce Racism. NYU Press.
• O’Neil, C. (2016). Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy. Crown.