Britain’s Illegal Betting Surge Puts Gambling Reform Under Scrutiny

The IFHA study examines 301 websites: 74 licensed and 227 illegal. Its relative traffic measure increased from 1.0 illegal-site visits per 100 licensed-site visits in August-December 2021 to 26.6 in 2025, before falling to 19.4 in January-July 2026.

Source: IFHA, Table 2. These are relative website-traffic figures for the sampled sites.

The underlying tables use sums of monthly unique-visitor counts. A person can appear in several months or visit several sites; these totals cannot be read as distinct annual customers.

Between the full years 2022 and 2025, sampled illegal traffic increased approximately fourteenfold while licensed traffic fell by 34%. Both sides of that movement affect the ratio.

A ratio of 19.4 to 100 also does not mean that 19.4% of gambling revenue went offshore. Website activity cannot establish how much visitors wagered, lost or withdrew. Nor can it show that a particular person switched operators.

The latest period therefore leaves two relevant findings: illegal traffic remained far above its early baseline, while its relative intensity was below the 2025 peak. January-July is also an incomplete year. Claims of uninterrupted growth throughout 2026 would overstate this evidence.

The financial history is more complicated because researchers have used different methods and definitions. Three frequently cited studies illustrate the problem.

PwC’s research, commissioned by the Betting and Gaming Council, also estimated that the number of online gamblers using unlicensed operators in the preceding year increased from approximately 210,000 to 460,000 between its two research rounds.

Frontier’s later BGC-commissioned study combined a survey of more than 6,000 people with operator and secondary data. Its £2.7bn figure concerned online activity. The larger £4.3bn estimate added in-person gambling, with the report acknowledging that respondents may have misclassified some activity under that broader definition.

H2’s May 2026 publication presents a much larger estimate for 2025 and says stakes more than tripled within its own 2019 to 2025 series. The BGC has used that research in its campaign for action against illegal operators. It is an industry research estimate, rather than a Gambling Commission measurement of the whole market.

These figures cannot be joined into a single line showing £1.4bn becoming £2.8bn, then £4.3bn and finally £16.6bn. Different researchers, coverage, geographies and assumptions stand behind them. In particular, the gap between Frontier’s 2024 estimate and H2’s 2025 estimate cannot establish a one-year growth rate.

All these monetary figures concern stakes. They are not interchangeable with customers’ net losses or operators’ gross gambling yield. Money won and wagered again contributes further turnover without requiring an equivalent fresh deposit.

The historical evidence supports concern about expansion. It does not yet provide one universally accepted financial measure against which every policy can be judged.

The government published its White Paper, High Stakes: Gambling Reform for the Digital Age, on 27 April 2023. Implementation followed through separate measures on different dates. Treating the White Paper as one restriction that suddenly took effect obscures that sequence.

Light-touch financial vulnerability checks began on 30 August 2024 at a threshold above £500 in deposits minus withdrawals over a rolling 30 days. The threshold fell to above £150 on 28 February 2025.

Those checks use public records to identify indicators such as bankruptcy or court judgments. They should be distinguished from more detailed financial risk assessments and from individual operators’ other customer checks. Describing the £150 threshold as a universal demand for bank statements would misstate the requirement.

Online slots then received statutory stake limits: £5 per game cycle from 9 April 2025, followed by a £2 limit for customers aged 18 to 24 from 21 May. These are slot restrictions; they are not limits on a horse-racing bet.

From 31 October 2025, operators also had to prompt customers to set a financial limit before their first deposit and make limits easy to review and amend.

Tax changes belong on a separate timeline. Remote Gaming Duty increased from 21% to 40% from 1 April 2026. That concerns online gaming, including casino products. The new 25% remote betting rate starts on 1 April 2027; remote bets on UK horseracing remain at 15%.

The distinction matters for the present argument. A betting duty change taking effect in April 2027 cannot explain traffic already observed by July 2026. The April 2026 gaming duty increase also cannot explain the earlier years of growth.

The sequence leaves a legitimate question about the cumulative effect of restrictions, customer processes and commercial decisions. It does not supply a causal answer. Changes in advertising, search visibility, payments, sporting calendars and enforcement could also affect activity.

A convincing assessment would examine comparable customers and products before and after a specific intervention, establish who was affected, and account for those other influences. Finding that traffic grew during a period of reform is a starting point for that work.

The BHA highlighted a separate follow-up covering 32 illegal websites taking bets on British racing. Their traffic relative to licensed sites rose from 0.97 per 100 in late 2021 to 6.8 in January-July 2026, approximately a sevenfold increase.

For racing, the concern extends beyond bookmakers losing business. Licensed betting supports the sport through the statutory levy, while illegal operators avoid the obligations attached to the regulated market. Betting outside that framework also creates concerns about integrity monitoring and consumer redress.

That gives the BHA a strong reason to press for action, alongside its institutional interest in racing’s finances. Neither point cancels the other. Its evidence deserves scrutiny, and its interpretation should be assessed with the same care as claims from operators, campaigners or the regulator.

The wider financial estimates cover gambling beyond horseracing. Applying a headline such as £16.6bn directly to racing, or treating it as the sport’s lost income, would create another misleading comparison.

The legal boundary is permission to serve Great Britain. An overseas operator can be properly licensed here; holding a licence elsewhere does not replace the required Gambling Commission permission.

The distinction between legal and illegal customers is less tidy. Banking research published by the Commission on 28 October identified 3,426 transactions involving 179 people through payment details linked to known or strongly suspected illegal operators. Among people with repeated detected illegal transactions, 91% also used licensed operators.

The study’s panel was not nationally representative, cryptocurrency was excluded and only part of illegal activity could be identified. The Commission explicitly says the analysis cannot establish the size of the illegal market or a causal relationship between illegal spending and harm.

We examined those findings in our coverage of Flutter’s response to the UK illegal gambling study. They complicate the idea that each customer simply leaves one market and enters the other. For some people, illegal gambling happens alongside continuing use of licensed accounts.

Our coverage of the Commission’s new Illegal Gambling Index explains another distinction: identifying participation answers a different question from estimating money wagered.

The strongest policy argument is that Britain needs a licensed market people can use confidently, supported by enforcement that makes illegal services harder to reach and fund.

Avoidable delays, confusing document requests and poor explanations should concern operators and regulators. A customer who cannot understand a legitimate check may be easier for an illegal competitor to attract. Improving that experience is compatible with maintaining effective protections.

The response must also account for people deliberately seeking to bypass exclusions or other safeguards. Making every restriction weaker would leave that risk unresolved. The purpose of a check, the friction it creates and the outcome for the customer all need examination.

Enforcement deserves the same attention to outcomes. Removing a domain has limited value if its audience and payment arrangements quickly reappear elsewhere. Useful measures would include sustained reductions in access, successful disruption of payment routes and evidence of improved consumer outcomes.

The Commission’s dashboard published on 5 November identified 586 accessible illegal websites in September. Its estimated engagement measure was 12% above the preceding three-month average but 11% below September 2025. Those observations concern a different monitoring system from the IFHA study, and they reinforce the need to specify the comparison behind a headline.

The government’s Illegal Gambling Taskforce already has a remit covering payments, online advertising and cooperation between enforcement bodies. That provides a practical framework for action against the services helping illegal operators find customers and receive money.

Britain should now demand clearer evidence of what those interventions achieve. The scale of the historical increase justifies urgency. The uncertainty around financial estimates makes transparency about methods more valuable, while the latest mixed traffic picture creates an opportunity to identify what may be working.

For racing, operators and consumers, the useful outcome would be fewer people exposed to illegal gambling and a stronger basis for judging which protections and enforcement measures deliver that result.