This week: research that asks children what they think, policy that has already decided what it thinks, and the timelines that do not add up.
🔍 This week I’ve been thinking about…
Apologies — it’s a long one…and yet more social media stuff.
On Monday evening, MPs voted 272 to 64 in favour of government amendments to the Children’s Wellbeing and Schools Bill that commit ministers to imposing social media restrictions on under-16s. Junior education minister Olivia Bailey told the Commons that, whatever the consultation outcome, the government would put age- or functionality-based restrictions in place, with curfews on top of rather than instead of those measures. Speaking to BBC Breakfast this morning, Education Secretary Bridget Phillipson confirmed the government’s intention to move before the end of the year.
Yesterday morning, I attended a webinar with the IRL Trial team.
The IRL Trial is (they believe) the largest scientific study of social media restrictions in healthy teenagers anywhere in the world. It is run jointly by Cambridge and the Born in Bradford team. The government has been softly promoting it alongside the consultation work.
Participants in Born in Bradford, a longitudinal cohort study tracking the lives of more than 60,000 Bradfordians, asked researchers in 2021 to look at whether social media was harming young people. This trial is what came out of that ask; timely, but not rushed in approach.
A few things from yesterday’s webinar are worth having on record.
The trial was co-designed with the young people who will take part in it. When the team asked them what they thought of a full ban, they said they would simply circumvent it. The trial therefore tests not a ban but a daily one or two hour cap on the social media platforms covered by the Australian ban, with a 9pm to 7am overnight curfew. Compliance is supported through an app with incentives such as leaderboards. Participants can override the time limits. This is, the team made clear, by design.
The team also made an ethical point that does not feature in any government messaging on this topic. Some young people, particularly those who are vulnerable or marginalised, rely on online communities for connection that is not available to them offline. Studying a full ban would mean removing that connection from research participants. Ethically, it’s just not ok to research.
The trial does not, and cannot, address use during school hours. Its research question is about end-of-day social media use. Schools are getting separate phone restrictions, by separate legislation, on a separate timeline.
The trial begins in September. Full analysis is expected by mid-2027.
Set those timelines against this morning’s announcement. The consultation that informs the policy closes on 26 May. The government has committed to laying regulations within twelve months, but is aiming to move before the end of this year. There is no mechanism by which IRL Trial findings can inform a regulatory framework that will be in place before the trial has produced any.
One further observation. The consultation was run by DSIT, the Department for Science, Innovation and Technology. This morning’s announcement, and yesterday’s parliamentary motion, came from DfE, the Department for Education. Unlike the phone ban during school hours and the new plans for digital/media literacy, it’s a bit confusing to me why DfE are now involved — and, bluntly, doesn’t feel like it bodes well for more holistic considerations about social media for everyone.
Changing the tone, slightly — along with the IRL Trial, we are repeatedly seeing surveys and data around how children are using — and cannot use — tech. This week, neurodiversity and edtech. The Digital Poverty Alliance report from earlier this month, covered below, reported that 90% of neurodivergent young people surveyed said the educational platforms they are required to use are not accessible.
The recurring question of this newsletter, expressed differently each week: who is the design for, what are they being asked, and whether their answers are allowed to go anywhere uncomfortable. The IRL Trial is doing the work to take that question seriously, and I’m really looking forward to the outcomes. The policy is not waiting for the answer.
📰 Three things worth your attention
1. Social media restrictions for under-16s even if no ban, minister says — BBC News
The political mechanics are worth following carefully. The Lords have voted four times in favour of opposition amendments calling for a faster ban; the government has now offered amendments that commit ministers to act, but on a longer timeline than the Lords had pushed for. The bill returns to the Lords for what looks like final consideration before royal assent. A progress report is due three months after that, with twelve months to lay regulations, though the government has signalled it intends to move before the end of the year.
The shape of the restrictions has not been decided. The decision to impose them has. Among the options on the table: a ban, age verification, functionality restrictions, curfews, or some combination. The consultation that closes on 26 May will shape the form, not the fact, of the action.
The political coalition supporting this is broad, and it includes parents who lost children in circumstances they associate with social media platforms. Their voices are essential to this conversation. The conversation also needs to include what we know, and what we are about to know, about whether the proposed mechanisms will deliver what they are being asked to deliver.
2. Digital Learning Dominates Schools But Remains Inaccessible to 90% of Digitally Excluded Young People — Digital Poverty Alliance / FE News
The DPA’s Accessibility in Digital Education Service Design Report is the clearest thing I have read this year on how educational technology fails ND learners by default. 88 respondents, primarily ADHD or autistic, the majority in Key Stages 3 and 4. The headline numbers are in the thinking section above; the report’s intersectional analysis is what gives it its weight.
ND learners who also have English as a second language navigate disorganised interfaces while relying on translation tools that struggle on those same interfaces. ND learners from ethnic minority backgrounds are statistically less likely to be diagnosed in the first place, so the support they would otherwise be entitled to does not arrive. ND learners in digital poverty access platforms designed for laptops on the only device they have, a smartphone. These exclusions don’t sit alongside each other. They land on the same children.
The recommendations are practical: interface customisation, flexible design, engagement with accessibility organisations including ADHD UK and ADHD Pirates, and inclusive consultation that genuinely reflects how accessibility intersects with other factors. Anyone choosing or designing software for schools should read it.
3. How the three largest US school districts are approaching AI in classrooms — Mashable
A useful primer on what district-level AI policy actually looks like in practice. NYC’s new guidelines use a “traffic light” approach, with explicit prohibitions on AI being used to create Individualised Education Plans, confer grades, make disciplinary decisions, or provide emotional and therapeutic counsel to students. LAUSD bans generative AI for under-13s and requires administrator approval for older students. Chicago has a detailed guidebook in which most chatbots, including ChatGPT and Claude, are currently not approved for student use.
Two things make this worth reading. It shows what consequential AI policy looks like when written by people who have actually thought about disability and child protection. And it ends on a sharp contrast: AI-only Alpha schools, private-equity-funded, replacing human teachers with screens for two hours a day.
A RAND survey quoted in the piece is also worth flagging. 80% of US students said their teachers had not taught them how to use AI for schoolwork. Fewer than half of school principals reported having an AI policy.
🔁 ICYMI
SWGfL Launches Survey to Gather Professional Views About Social Media Banning — UK Safer Internet Centre
SWGfL has launched a survey of professionals working with children, asking what they think a possible under-16 social media ban should look like. This is, usefully, the inverse population to the government’s own consultation, which asks children, parents, and the public. The professionals being asked here are the ones who will be doing the actual work of supporting children through whatever policy lands. If you work in this space, fill it in.
🔬 What’s new with CAISE
Forms forms forms! The logistics of empirical research are…not small. This week, we’re aiming to get as many ducks in a row as we can so that we can get started asap!
→ What are you seeing in your school, your research, or your own use of AI this week?
Let me know, or share this with someone who is trying to figure it out.