Fewer young people in Britain report spending their own money on gambling, according to the Gambling Commission’s latest annual survey. The decline is welcome, although the research found no statistically significant change in the proportion meeting the threshold on its youth-adapted problem gambling screen.
The Young People and Gambling 2026 report, published on 12 November, found that 26% of 11–17-year-olds had spent their own money on a gambling activity in the preceding year, compared with 30% in 2025.
Ipsos surveyed 3,867 pupils in England, Scotland and Wales. The findings cover a broad range of activities, including those legally available to children, and deserve a more careful reading than a single headline about underage betting.
For operators, regulators and families, the central questions remain distinct: what children participate in, whether they can access age-restricted products, and whether they experience adverse consequences.
What the participation figure includes
The largest individual activity was spending money on arcade gaming machines, such as penny pushers and claw grabbers, reported by 19%. Betting between friends or family accounted for 11%, while 5% reported playing cards for money.
These categories overlap: a pupil could report more than one activity. They cannot be added together to produce a separate total.
| Measure in the 2026 survey | Reported proportion |
|---|---|
| Spent own money on any gambling activity in the previous 12 months | 26% |
| Spent own money on non-age-restricted gambling activities | 24% |
| Spent own money on age-restricted gambling activities | 5% |
| Scored four or more on the youth-adapted problem gambling screen | 0.9% |
Percentages use the full survey population as their base. The measures overlap and should not be summed.
The age-restricted figure is particularly relevant to access controls. It also shows why the overall 26% should never be presented as the proportion of children using online casinos.
At the same time, an activity being legally accessible does not establish that every child’s experience of it is harmless. Legal status, participation and consequences answer different questions.
Progress has been uneven
The recent history provides a useful check on any claim of a smooth decline. Own-money participation stood at 27% in 2024, rose to 30% in 2025 and returned to 26% in 2026.
The longer series starts at 31% in 2022. However, coverage expanded in 2023 to include Year 12 pupils and independent schools, so the Commission cautions that comparisons with 2022 are indicative.
That leaves a measured interpretation: the latest estimate is lower, following a rise the previous year. The sequence cannot identify which policy, commercial change or wider social influence produced the movement.
In our view, the industry should welcome the direction while resisting the temptation to turn one annual result into a verdict on the success of its safeguards. Establishing that success requires evidence about the controls themselves and the children they are intended to protect.
A harm measure needs a separate reading
The survey’s DSM-IV-MR-J screening measure asks about reported behaviours and experiences associated with gambling difficulties. A score of four or more met its problem gambling threshold for 0.9% of young people, against 1.2% in 2025.
The Commission found no statistically significant difference between those estimates. The smaller point estimate therefore does not establish a reduction in the underlying rate. Equally, the absence of a significant difference does not prove that nothing changed.
This is a screening classification based on self-report, rather than a clinical diagnosis or a measure of addiction. Children within the same category may report different behaviours and levels of difficulty.
The distinction matters when choosing a response. Reducing occasional participation across a large population and helping a smaller group experiencing difficulties are related objectives, but progress on the first does not automatically demonstrate progress on the second.
A useful assessment needs to retain both measures. Otherwise, an improving participation headline could obscure the outcomes that most urgently require attention.
The existing safeguards still need testing
Britain has already strengthened the barriers to adult gambling products. Changes introduced in May 2019 required affected online operators to complete customer verification before allowing gambling.
The current remote social responsibility code requires age verification before customers can deposit funds, access operators’ free-to-play gambling games or gamble using money, free bets or bonuses. It also requires licensees to monitor the effectiveness of their underage gambling prevention procedures.
The survey cannot establish which licensed business, account or access route sits behind an individual response. It would be unjustified to treat every reported instance as a verified breach by an identifiable operator.
Those limits nevertheless leave a practical question for the industry: how do reported experiences compare with compliance evidence, account investigations and other research? The appropriate response is to investigate that connection and test where protections work or fail.
The Commission’s research and statistics director, Ben Haden, identified preventing children from spending money on age-restricted products as the regulator’s focus when announcing the findings.
Advertising reaches beyond verified accounts
The report also found that 89% of young people recalled seeing or hearing gambling advertising or promotions during the previous year. Recall measures exposure as respondents remember it; it does not demonstrate that a particular advert caused a child to gamble.
Advertising controls have their own history. Since October 2022, the CAP Code has applied a stronger test concerning gambling advertisements’ appeal to under-18s.
A concrete example came on 27 May 2026, when the Advertising Standards Authority upheld a complaint about Oddschecker Instagram posts featuring Harry Kane and Erling Haaland. The ASA regarded the posts as marketing and concluded that the footballers had strong appeal to under-18s. It also found that the account’s adult audience restrictions did not robustly exclude children.
That ruling illustrates the difference between verifying a gambling customer and controlling who encounters promotional content. A secure registration process cannot, by itself, answer every question about a brand’s wider advertising audience.
Responding to the survey, Ygam chief executive Emily Tofield welcomed lower participation while warning about the normalisation of gambling and the overlap with gaming, according to NEXT.io. That is a relevant concern to examine, while keeping claims about causes separate from the survey’s observations.
Loot boxes require their own interpretation
Across all respondents, 22% reported opening virtual loot boxes or surprise boxes using their own money. The report examines this through its gaming questions; the figure should not be added to gambling participation or read as proof that all such purchases constitute illegal gambling.
GambleBallers recently examined planned research into children’s use of paid random rewards, involving Lancaster and Edge Hill universities. That project aims to investigate how engagement develops over time; it has not yet delivered completed findings.
The new survey supplies a picture of reported activity. Following children over time can address a different question: what changes first, and how purchasing relates to later experiences. Both approaches can contribute without being treated as interchangeable evidence.
What the survey can establish
Fieldwork ran from 12 January to 12 June 2026, with pupils completing questionnaires in class. November’s publication date therefore should not be mistaken for a measurement of behaviour through November.
The Commission sets out limitations involving recall, interpretation and disclosure, alongside its quality checks. The sample also excludes some young people, including those outside mainstream education or absent when the survey was conducted.
One further change deserves attention: the 2026 research was a standalone study, whereas earlier waves formed part of a broader omnibus questionnaire. The Commission reports no evidence that this materially affected the findings, while advising that the difference be considered in comparisons.
Our view is that the report is most useful as a guide to further scrutiny. Lower participation is encouraging. Establishing whether children are better protected requires the access figures, screening results and advertising evidence to be examined together, alongside other sources.
The next test is whether those findings lead to demonstrably effective protections and better outcomes for the young people who still report difficulties.