Mobicip interview series for Safer Internet Day 2026 · Part 6 of 8
Child Safety by Design
Watch the recording on YouTube
Summary
Anitha asks whether today’s tech leaders are failing to prioritize child safety, what would get entrepreneurs to design for it, and how parents, policymakers and educators can push the industry.
I say we’re at a critical junction. In the next couple of years, probably two, the way models and applications are designed will become rigid and difficult to change. This is the moment for companies to work with children’s organizations to understand how children actually use their technology.
Key points
- Companies need to admit that children under their published minimum age are using their products, sometimes in large numbers. I point to regulators such as Ofcom in the UK, the eSafety regulator in Australia and the EU Commission, which are now pressing companies to find out who the children on their platforms are.
- Safety education has to fit a child’s age and stage. You’d teach internet safety or AI safety very differently to a three-year-old, a ten-year-old and a fifteen-year-old.
- The first major project of AIChildSafety.org is the Childhood & AI Lab Research Fellowship. For my whole career in technology and social issues I’ve faced a shortage of evidence about what works, and I want research that measures how children respond, what harms we’re seeing, and how technology and AI can help.
Speakers
- John Zoltner Guest
- Anitha Swaminathan Interviewer, Mobicip
Transcript
The wording is as spoken, with punctuation and obvious errors corrected. See “How this transcript was made” above.
[00:00:00]Anitha Swaminathan: So do you think today’s tech leaders are failing to prioritize child safety, and what would it take for entrepreneurs and innovators to prioritize safety by design, and how can parents and policymakers and educators push the industry in that direction?
[00:00:26]John Zoltner: Well, I’m looking especially at how AI can affect child safety, and as you know, we’re at the very beginning of the stage of AI development and the way that children will use AI. So to me it’s we’re at a really critical junction where in the next couple of years, really probably two years, the way the models and the applications are designed will really grow, set, will be rigid, and it will be difficult to change the way they’re designed after that time. So I just think it’s a critical moment for the companies to be working with children and children’s organizations that represent or that provide safeguarding services for children to really understand how children are using their technology, and that includes, you know, if they’re using it in ways that the companies didn’t envisage them using it. They and they need to admit, recognize and admit that children who are under the minimum age that the company publishes that is allowed to use their technology, there are children under that age that are using the tech, and many times many children, there could be a fairly large percentage of users who are under the publicized age for using their networks. So we need to get better at understanding who on platforms are children, and I think companies are taking that more seriously because regulators like Ofcom and the UK and the eSafety Commission in Australia and the EU Commission are requiring them to do that, and they’re finding them if they’re not keeping children off of their platforms, and they, and we need to also, we need to have training that is appropriate for children at the ages and stages of development where they’re at. So you teach internet safety very differently, or AI safety very differently, to a three-year-old and a 10-year-old and a 15-year-old, and we need to be investing in, yeah, in the understanding what works for children of all those ages. So the first big project under AIChildSafety.org that we rolled out is we’re creating a fellowship for the Childhood and AI Lab Research Fellows program. We did a call for fellows about a month and a half ago, we had a tremendous response, so we’re still combing through applicants to be fellows in that project, and in mid-January [2026] we’re going to announce our first cohort of fellows. And the reason I made that our first major project is because for my entire career in technology and social issues I’ve been facing the problem of not having enough research and not having evidence to be able to make decisions based on research about what we do with technology. And it’s the same with online protection, you see these sometimes protection campaigns, behavior change campaigns that obviously aren’t going to work, and they’re things that were thought up by marketing teams but without participation of children, or thought up by the organizations that are developing them. And I want to see real research that shows what works with children and that measures, the way an advertising or marketing company does very minute measurement of their audience’s responses to their message. So I want to understand what children react to, I want to understand what the harms we’re seeing from use of AI or use of technology in general by children are, and I want to see research that shows how we can respond to those harms, and how we can use technology or AI to help children in a positive way.