Maybe The Edtech Backlash Is a Fight Over Defaults?
Last weeks AFT plan caught my attention because it showed where the argument has moved. The headline issue is now what should be switched off by default for younger children, and what should need a very deliberate decision to switch back on. AFTs announcement and Education Weeks report make that shift very clear.
The Gallup survey gives the awkward middle of the story. Just 18% of U.S. adults working in K-12 settings receive formal guidance on how AI tools should be used at work, and about a third receive no guidance at all. That is a huge gap between adoption and governance. Formal guidance is still rare, which leaves every new tool to be interpreted locally. Gallups findings are useful because they show how quickly policy can lag behind use.
I also noticed the call for safety research and training that sit outside the companies building the tools. That feels like one of the more practical ideas in the whole debate. If the same vendor is shaping the product, the training, and the safety story, scrutiny gets thin very quickly. EdWeeks reporting covers that part well, and it is worth reading in full alongside the AFT release.
The wider screen-time backlash matters too. AP reports that the pandemic accelerated device rollout, that classrooms became saturated with screens, and that parents and adults around children are now pushing back on how much of the day disappears into devices. APs report describes the shift well: after the rush to put a device in every child's hands, the conversation is now about whether those defaults were too generous, too early, or too hard to reverse.
From a software design point of view, that is the useful question. If a child-facing product needs a training deck, a policy memo, and a lot of local judgement just to stay on the rails, the architecture is doing too little work. We keep coming back to the same things in our own product thinking: age gating, data minimisation, auditability, reversible settings, and communication models that responsible adults can understand quickly. That is a product question, and a procurement question, because default settings become policy in practice.
That is why we think in terms of architecture first. In an open world for children, the defaults matter more than the marketing. Communication should be constrained by design. Cross-group connection should require an active adult decision. Currency should be earned, not bought. Trading should be absent. All of that reduces exposure and changes the shape of the risk before the first child arrives.
I do not have a settled view on the right national policy here, and I do not think software teams should pretend to be classroom experts. But I do think this moment is forcing a better question on the whole industry: what would child-facing software look like if safety, transparency, and reversibility were the starting point rather than the patch?
If you work with children, build for them, or spend time thinking about their digital lives, I would be interested in your view.
We’d love to hear your thoughts on this. Find us on the social channels linked at the top of the page.