Gambling Commission
The Gambling Commission published a new Illegal Gambling Index on 5 November, using seven reported behaviours to identify likely engagement with illegal websites. Released alongside an updated monitoring dashboard, it gives Great Britain's gambling debate a new source of consumer evidence. 1
The focus is on websites offering gambling to consumers in Great Britain without the required Commission licence. A licence elsewhere does not replace that permission. 5
The first estimates classify 0.6% of adults in the past-four-weeks category. Among people who gambled online in the previous year, excluding National Lottery-only players, the corresponding estimate is 2.5%. 2 The denominator changes the headline substantially. Neither figure describes the proportion of gambling revenue going to illegal operators.
That distinction is central to the news. Claims about the black market influence arguments over taxes, customer checks and the responsibilities of licensed businesses. A better measure can help test those claims, provided the conclusions stay within what the evidence can establish.
Why Asking People About Illegal Gambling Was Not Enough
The index addresses a basic problem: consumers may not know whether the website they used was licensed. In the piloted questions, some respondents who named websites they believed were illegal actually identified licensed operators. The Commission consequently excluded the direct question about intentional illegal use from the index calculation. 3
Instead, the measure draws on reported experiences: knowingly using a site that does not participate in Gamstop; choosing one to avoid age or identity verification; reversing a withdrawal; using a VPN to reach an otherwise unavailable gambling site; and gambling with cryptocurrency, NFTs or virtual assets, or money directly from a credit card. 2
Index methodology
Technical report: Gambling Commission and Yonder Consulting.3
These experiences carry different weights. The technical report gives credit-card use a weaker rating, partly because respondents may confuse credit and debit cards. Other indicators carry greater weight, and older recollections contribute less to the score. 3
The approach is more thoughtful than treating every affirmative answer as equally convincing. It also retains a judgement about what each answer means. A person can misunderstand a payment method, a website feature or when something happened. Combining indicators helps organise that uncertainty; it cannot remove it.
The resulting categories describe survey respondents' likely engagement. They should not be used as a legal ruling on a named operator or as proof that every reported experience occurred exactly as recalled.
Reading the First Figures Properly
The findings cover four survey waves from April 2025 to April 2026. The online questionnaire's eligible online-gambling sample comprised 3,569 respondents. 2
| Population used for the estimate | Past-four-weeks classification | Additional past-12-months classification |
|---|---|---|
| Adults aged 18 and over in Great Britain | 0.6% | 0.4% |
| People who gambled online in the past year, excluding National Lottery-only players | 2.5% | 1.5% |
These are weighted estimates. The two index categories are mutually exclusive; the second excludes the first. A further 4.6% of adults showed limited indications, with insufficient evidence for classification into either use category. 2
That last group must remain separate. Adding it to the classified users would replace the research's confidence distinction with a much broader definition of illegal participation.
There is an equally important financial distinction. A participation estimate counts people, while a market-share estimate requires information about their gambling activity or expenditure. Customers spend different amounts and may use both licensed and illegal services. The percentage of people identified by this index cannot be read directly as a percentage of stakes, losses or operator revenue.
Publication timing also matters. November's release describes survey responses collected up to April. It cannot establish what happened to participation later in 2026.
A Debate with a Long History of Competing Evidence
The argument predates this index. Parliament's Culture, Media and Sport Committee examined it in its 2023 gambling report. The committee recorded industry warnings that intrusive regulation could push customers towards unlicensed operators, alongside challenges to the scale of that threat. It called for better understanding and monitoring, while concluding that the risk should not deter appropriate regulation of licensed gambling. 4
The Commission subsequently developed several ways to investigate the problem. Its October 2024 methodology paper combined website identification and traffic estimates with assumptions drawn from licensed gambling behaviour. That paper explicitly acknowledged incomplete coverage and limitations in transferring those assumptions to the illegal market. 5
More recent research has added different perspectives. Phase 3, published on 20 October 2026, explored access restrictions, demands for personal information and differing consumer attitudes. It deliberately focused on illegal-site users to understand their motivations, rather than measuring participation across everyone who gambles. 6
Banking analysis published on 28 October examined detected payments and found that illegal transactions often occurred alongside licensed gambling. 7 We explored that evidence in our coverage of Flutter's response to the illegal gambling study.
Our Flutter Entertainment overview sets that response in the context of the wider corporate group.
The new index adds a survey measure to that developing evidence base. These methods answer different questions. Banking records can reveal identifiable payments; website monitoring can indicate visible activity; surveys can capture experiences and characteristics. Agreement between them would strengthen an interpretation. Differences need explanation before anyone presents a single number as the whole market.
The Tax Argument Still Needs a Causal Test
There is a plausible commercial argument that making a licensed product less attractive could encourage some customers to consider alternatives. Establishing how often that happens, and what caused it, requires more evidence than finding that someone has used an illegal website.
To assess a particular tax change or customer check, researchers would need comparable observations before and after it, an account of who was affected, and a way to distinguish its effects from other influences. Changes in promotions, sporting events, enforcement and website availability could also affect observed behaviour.
An account of frustration is useful evidence about a customer's experience. It does not automatically identify the effect of a specific rule. Likewise, a low population estimate cannot demonstrate that every protection works well or that the people who do use illegal services face little risk.
The dashboard provides a good example of why comparisons need care. Estimated engagement time on identified illegal sites was 12% higher in September 2026 than the previous three-month average, but 11% lower than September 2025. 8 Either direction could supply a headline; both describe the same month against different baselines.
That engagement measure combines modelled visits and visit duration. It has coverage limits and includes an adjustment for VPN use. It is separate from the survey index and does not directly measure money wagered. 8
Evidence of a short-term increase therefore needs its longer context. A year-on-year decline also needs checking for changes in what the monitoring system can see. Neither observation, on its own, identifies a tax effect.
Our View: The Same Evidential Standard Should Apply to Everyone
We welcome the index because it makes the assumptions behind a headline more visible. Its greatest contribution could be a consistent way to ask who is engaging with illegal gambling and how that changes over time.
Licensed operators have legitimate reasons to care about unlicensed competition. Their claims about the effects of regulation should specify the behaviour being measured, the period involved and the basis for attributing a change to policy. The regulator should make its own claims with the same precision.
Consistency will be particularly important as the method develops. The Commission says indicators and scoring may change as evidence improves. 1 When that happens, published comparisons should explain whether earlier results have been recalculated or whether the change creates a break in the series. Otherwise, a methodological revision could be mistaken for a change in consumers' behaviour.
The practical response should also distinguish between customers' needs. Clearer explanations and better service can address avoidable frustration. People seeking to bypass an exclusion require a different protective response. Treating every route into illegal gambling as the same problem would make both research and policy less useful.
The index supplies a stronger starting point for that work. The next advance will be evidence that connects measured changes to specific causes, then shows whether the response improves outcomes. Britain's black market debate now has a more structured measure; the claims built around it still have to be earned.
Prepared with help from ChatGPT.