Online safety · Analysis
Ofcom’s new survey, scaled up with today’s ONS population figures, puts about 1.9 million UK 11 to 17-year-olds in contact with people they do not know. The under-16 ban reaches around 1.3 million of them. The rest depend on a default setting.
THE QUESTION
How many children does the ban actually reach?
Ofcom’s latest study of how children communicate online found that 30% of 11 to 17-year-olds who use social or messaging apps talk to a mix of people they know and people they don’t. Another 6% are connected mainly with strangers. A third would reply to a message from someone they don’t know who added them.
A percentage is easy to read past. A headcount is harder to ignore, so we ran Ofcom’s figures against the Office for National Statistics’ mid-2025 population estimates, which came out on 1 October 2026.
The government wants its under-16 social media ban in front of Parliament by the end of 2026. Ofcom has until the end of October to tell MPs how a platform can prove someone is over 16. So the useful question now is how many of these children the ban actually reaches, and how many it leaves to a settings menu.
Key findings
- About 1.9 million UK 11 to 17-year-olds are connected with people they don’t know personally. The maths ↓
- The share is already 29% to 32% at ages 11 to 15, rising to 43% at 16. By age ↓
- 34% of children on social or messaging apps would reply to a stranger who added them, against 23% in Ofcom’s 2023 survey (an indicative comparison). Then and now ↓
- The under-16 ban covers 1.29 million of them; 0.65 million aged 16 and 17 rely on a default setting, and it all depends on an age check Ofcom defines by the end of October. The ban ↓
THE SURVEY
One in three children have strangers in their contacts
Ofcom’s research , carried out by Savanta, surveyed 4,097 children aged 11 to 17 between 13 February and 19 March 2026. Results are weighted to match the UK population by age, gender, nation, region, ethnicity and socio-economic grade.
Of the children who use a social or messaging app, 64% say they are connected mainly with people they know personally. 30% have a mix of known and unknown contacts, and 6% are connected mainly with people they don’t know. Two in three get friend requests from people they have never spoken to at least weekly, and 14% get them every day.
How do they decide who to accept? 39% would accept a request because of mutual friends online, and 27% based on the person’s age. A profile’s age proves very little, as our report on AI fake identities and sextortion scams sets out.
This isn’t just older teenagers. Our breakdown by single year of age shows the share sits between 29% and 32% from age 11 to 15, then jumps to 43% at 16.
Children connected with people they don’t know, by age
Estimated UK headcount: Ofcom weighted share for each age, applied to the ONS mid-2025 population for that age.
Orange: under 16, covered by the planned ban. Grey: 16 and 17, protected only by default settings.
Chart: World of Chat · worldofchat.co.uk
Download the age-by-age data (CSV)THE CALCULATION
The real headcount is 1.9 million, not 1.7 million
The quick way to do this sum is to take 30% of every UK 11 to 17-year-old. That gives 1.75 million, but it is wrong in two ways that cancel out only partly.
First, Ofcom’s percentages are based only on children who use a social or messaging app, which is 93.7% of the weighted sample. Second, the 30% leaves out the 6% who are connected mainly with strangers. Fix both, work it out age by age, and the number goes up.
Headcount = sum over ages 11 to 17 of (ONS population × Ofcom weighted share connected with unknown people)
Ofcom weighted counts: 1,144 “a mixture” + 216 “mainly people I don’t know” = 1,360 of 4,097 children, or 33.2% of all 11 to 17s. Summed by single year of age against the ONS total of 5,831,252, that gives 1,935,876 children. Under-16s (ages 11 to 15): 1,285,255. Ages 16 and 17: 650,621.
For comparison, Ofcom’s 30% applied to app users alone gives 5,831,252 × 93.7% × 30% = 1.64 million.
From all UK teenagers to the ones talking to strangers
UK children aged 11 to 17, mid-2025. Headcounts derived from Ofcom weighted shares.
Chart: World of Chat · worldofchat.co.uk
The figure is probably an undercount. Ofcom left gaming platforms out of the survey, and its separate children’s media research from May 2026 found that 55% of 8 to 17-year-olds who game online play with people they have only met online at least some of the time.
THEN AND NOW
Replying to strangers looks more common than in 2023
Ofcom ran a similar survey in December 2022. Back then, 23% of 11 to 18-year-olds said they would be likely to reply to a message from someone they didn’t know personally. In 2026 it is 34% of 11 to 17-year-olds who use these apps.
Ofcom warns that the platform list, the question wording and the age range all changed, so the comparison is indicative rather than a measured trend. Even allowing for that, the gap on replying is wide.
Ofcom’s 2023 and 2026 surveys compared
Percentage of children. Indicative only: methods and age range differ between waves.
Would reply to a stranger (%)
Connected with people they don’t know (%)
Chart: World of Chat · worldofchat.co.uk
Willingness to reply peaks at 17, where 42% of app users say they are likely to answer a stranger. That is the age group the ban leaves on default settings.
WHAT CHANGES
What the ban changes, and for whom
On 15 June 2026 the government said social media firms will be barred from offering services to under-16s, using Australia’s model. Snapchat, TikTok, YouTube, Instagram, Facebook and X are named. It also promised to block livestreaming and stranger contact for under-16s on a wider range of services, including games.
The powers come from the Children’s Wellbeing and Schools Act 2026 , which lets ministers act through regulations. Both Houses must approve them. The government’s fact sheet says the first regulations will be laid before the end of the year, with changes in spring 2027.
| Group | Connected with strangers | What changes in spring 2027 |
|---|---|---|
| Under 16, on social media | 1.29m | Blocked from covered platforms such as Snapchat, TikTok and Instagram |
| Under 16, on games and other services | Not measured | Stranger contact and livestreaming restricted; exact services still to be listed |
| Under 16, on WhatsApp and similar | Not split out | Messaging apps are not intended to be part of the social media ban |
| Aged 16 and 17 | 0.65m | Stranger contact and livestreaming off by default, and a midnight to 6am curfew on by default. Teens can switch these defaults off |
WhatsApp is the gap to watch. Ofcom found 74% of children use it, more than any other app in the survey. The government says under-16s will still be able to stay in touch with “known friends and family” on messaging services. It has not yet published the full list of services the stranger-contact restriction will cover.
AGE ASSURANCE
Everything rests on the age check
A ban only reaches the 1.3 million if platforms can tell who is under 16. In July, Ofcom’s report on age assurance found age checks spreading fast but warned that the methods many social media firms rely on would not hold up. It opened a formal investigation into TikTok’s approach on the same day.
“…current age inference methods are insufficient in preventing children from accessing a service…” Ofcom, on social media firms, press release, 15 July 2026
Age inference means guessing age from behaviour and signals. The alternative is a check at sign-up, such as photo ID matching or facial age estimation. For now these checks fall on the big platforms, and they cost real money. We’ve worked out what they would cost a small site in our piece on Online Safety Act age verification costs . Ofcom’s end-of-October assessment will set out what “highly effective” looks like at 16 rather than 18. MPs get it before they debate the regulations.
Two leading UK child safety charities split on whether a ban works at all, and both point to Australia.
“Australia has shown that bans alone don’t keep all children safe.” Chris Sherwood, chief executive, NSPCC, 15 June 2026
The NSPCC welcomed the ban, with conditions: robust age checks and real enforcement. The Molly Rose Foundation, set up by the family of Molly Russell, argued against it two days before the announcement.
“In Australia this is not working, it’s not something that is enforceable.” Andy Burrows, chief executive, Molly Rose Foundation, to BBC Breakfast, 13 June 2026
Some platforms are moving ahead of the law. In May, Snap agreed to stop adult strangers contacting UK children by default and to stop encouraging children to add people they don’t know. That is the same stream of requests two in three children in Ofcom’s survey get every week.
The other way this ends is a shutdown. Omegle paired strangers at random, children included, and closed in 2023. Our piece on why Omegle had to end covers what went wrong.
PRACTICAL NEXT STEPS
What to watch, and what to do now
The ban could cut stranger contact for about two-thirds of the children who have it. The other third are 16 and 17, and they keep their accounts with a safer default they can switch off. Whether the under-16 part works depends on an age check Ofcom has not yet defined.
If you are a parent of an under-16
Don’t wait for spring 2027. Odds are about one in three that your child already has strangers in their contacts, so set accounts to friends-only now and ask who they actually know.
If your child is 16 or 17
The new rules only change the default. Agree with them now whether stranger contact stays off, because 42% of 17-year-olds on these apps say they would reply.
If you follow the policy
Read Ofcom’s October age assurance report first. If it doesn’t rule out age inference at 16, the 1.3 million figure is a ceiling the ban won’t reach.
If your child mainly games
This survey did not cover games. Check in-game chat and voice settings yourself, since the government has not yet listed which gaming services the stranger-contact rule will cover.
REFERENCES
Sources
Cite this analysis
World of Chat (2026). “1.9m UK Kids Chat With Strangers. The Ban Changes That.” Analysis of Ofcom and ONS data, October 2026. https://www.worldofchat.co.uk/uk-kids-chat-strangers-ban/
Free to quote and republish the figures and charts with a link to this page. Underlying data: Ofcom and ONS .
- Ofcom and Savanta, “Understanding Online Communications Among Children” research report (2026, page updated 9 September 2026): 30% mixed contacts, 6% mainly unknown, 34% would reply, 64% weekly requests, 39% and 27% acceptance reasons, 74% WhatsApp use, sample 4,097, fieldwork 13 February to 19 March 2026.
- Ofcom, data tables for the 2026 study : Tables 4, 28 and 45, weighted counts by single year of age.
- Ofcom, 2026 technical report : recruitment, weighting and fieldwork dates.
- Ofcom, “Understanding online communications among children: quantitative research” (November 2023, fieldwork December 2022): 23% would reply, 32% connected with people they don’t know.
- ONS, “Population estimates for the UK, England, Wales, Scotland and Northern Ireland: mid-2025” (1 October 2026): table MYE2, UK persons aged 11 to 17.
- GOV.UK, “Social media to be banned for under-16s” (15 June 2026): scope, named platforms, stranger contact and livestreaming restrictions.
- GOV.UK, “Fact sheet: New rules to protect children online” (updated 18 September 2026): timetable, messaging exemption, 16 and 17 defaults, curfew.
- GOV.UK, “Growing up in the online world: government response (July 2026)” : midnight to 6am curfew applied by default, which 16 and 17-year-olds can turn off.
- GOV.UK, “Growing up in the online world: progress statement” (2 July 2026): regulation timetable and affirmative procedure.
- Children’s Wellbeing and Schools Act 2026 : powers to make age restriction regulations.
- Ofcom, “Use of Age Assurance Report 2026” (July 2026) and accompanying press release (15 July 2026): age inference warning, TikTok investigation, end-of-October over-16 assessment.
- Ofcom, “Tech firms commit to stronger anti-grooming measures” (May 2026): Snap default protections.
- Ofcom, children’s media lives research (May 2026): 55% of young gamers play with people met only online.
- NSPCC, “Social media ban is a win for children and parents” (15 June 2026): Chris Sherwood quote.
- BBC News, “Social media ban for under-16s would be unenforceable, charity warns” (13 June 2026): Andy Burrows quote.
Calculations are World of Chat’s own, derived from the cited figures. Headcounts apply Ofcom’s weighted share for each single year of age, with all respondents as the base, to the ONS mid-2025 UK population for that age, then sum. “Connected with strangers” combines Ofcom’s “a mixture” and “mainly people I don’t know personally” answers. “Would reply” is the “fairly likely” plus “very likely” share for someone who added them and whom they don’t know personally. Survey figures carry sampling error, and single-year-of-age estimates rest on bases of roughly 480 to 770 children, so they are rounded to the nearest thousand. The 2023 and 2026 comparisons are indicative only, as Ofcom advises. Last checked: 1 October 2026.