Twenty-five million people do not pick up a habit overnight, nor do they abandon it because a president signs a piece of paper. Yet on September 25, Luiz Inácio Lula da Silva did exactly that, staking his political capital on the belief that Brazil’s online gambling explosion had become a machine for converting human weakness into balance-sheet growth. His Provisional Measure 1,394 gave licensed operators six days to go dark, froze R$36.9 billion in annual turnover, and told 25 million registered bettors to empty their accounts by midnight on October 5. The block takes effect at 00:00 on October 6, two days before Brazilians vote in general elections. Lula told the UN General Assembly the industry was “transforming addiction into profit.” The phrasing was deliberate, moral, and economically brutal. It resonated in a country where household debt among the poorest has become a national emergency, and where gambling advertisements follow people from township streets to taxi screens to WhatsApp groups. The question for South Africa is not whether Lula chose wisely, but what happens when a government decides the social cost has finally exceeded the tax revenue. South Africa must also consider if there is any middle ground between open season and scorched earth.
The Market That Grew Too Fast
Brazil’s regulated betting framework was barely a year old when Lula killed it. In its first twelve months, the market generated R$36.9 billion, roughly €6.2 billion at prevailing rates. The federal government had begun collecting its share aggressively. From January to July 2026, operators paid R$8.7 billion in tax, up 76.86% from the R$4.9 billion collected in the same period of 2025. The full-year 2025 total was R$9.95 billion. Part of the jump came from a GGR tax rate increase, from 12% to 13%. Most came from volume. Around 25 million Brazilians held active accounts with licensed operators. The seventeen largest gambling domains logged approximately 1.1 billion visits in the thirty days before the ban.
Betano, operated by Kaizen Gaming, dominated. The platform recorded 334 million visits in that final month, roughly 30% of tracked traffic and an estimated 23% of 2025 market revenue by year-end. Superbet, backed by Super Group, drew 175 million visits. bet365, the British-listed giant, logged 101 million. No other brand cracked nine figures. The top three accounted for 56% of all visits. This was not a fragmented market; it was concentrated, with global operators who had invested heavily in local infrastructure, Portuguese-language customer service, and celebrity sponsorships that turned football commentators into walking billboards.
The financial commitments were substantial. A federal licence cost R$30 million, approximately €5.1 million. Those licences now lapse on October 25 without refund. Entain, the FTSE-listed group behind Ladbrokes and Coral, has already cut its 2026 online NGR growth guidance to 4-6% and warned EBITDA would hit the low end of its range. FIRST.bet, a supplier to Brazilian sportsbook clients, disclosed those relationships generate over €350 million annually in GGR. The supply chain damage extends well beyond the operators whose logos appear on television.
The Justification and Its Numbers
Lula’s government did not rely on moral argument alone. It cited annual social costs of R$38.8 billion, slightly above the market’s gross revenue. It pointed to 1.3 million self-exclusion requests, formal admissions by registered users that they could not control their own behaviour. The self-exclusion figure is particularly telling because it represents only the minority who recognised their problem and navigated the bureaucratic process to do something about it. It excludes the unknown millions who never reached that point, or who tried and failed, or whose families discovered debts rather than habits.
The ban’s mechanics are uncompromising. Deposits stopped immediately on September 25. Players must withdraw existing balances by 23:59 on October 5. From midnight the following day, internet service providers will block licensed sites. The 120-day congressional review window for provisional measures means the ban could theoretically lapse if legislators reject it, but the immediate destruction is already complete. Operators have begun delisting apps, terminating local staff, and redirecting marketing spend to markets where their presence remains legal.
The legal pushback is equally swift. The National Association of Games and Lotteries, ANJL, has asked the Supreme Court to suspend Provisional Measure 1,394. bet365 has publicly called the ban unconstitutional and thrown its support behind trade body challenges. The constitutional argument centres on whether a provisional measure, designed for urgent and exceptional situations, can properly be used to annihilate an entire legal industry that the same government created by statute. The Supreme Court’s decision will shape whether Brazil’s gambling market stays dead or staggers back to life in some diminished form.
The South African Parallel
South Africa has not banned online gambling. It has not come close. The National Gambling Act of 2004 predates the smartphone revolution, and regulatory updates have struggled to keep pace with offshore operators who accept rands, sponsor local sports teams, and advertise on platforms the Advertising Regulatory Board cannot effectively police. The result is a grey market that functions as an open one, with licensed domestic operators competing against unlicensed international platforms that face no meaningful enforcement.
The advertising is impossible to miss. Gambling promotions dominate commercial breaks during football broadcasts, appear on billboards in urban townships, and circulate through influencer marketing that presents betting as a lifestyle choice rather than a probability exercise. The messaging targets young men disproportionately, employs sports celebrities to normalise daily wagering, and frames small-stake bets as harmless entertainment. For households with disposable income and financial literacy, this may hold true. For households already carrying unsustainable debt, the pitch lands differently.
The debt picture is well-documented. The National Credit Regulator has repeatedly flagged rising unsecured lending to lower-income households, with debt-to-income ratios that leave no margin for loss. The South African Reserve Bank’s quarterly data shows consumer credit extension growing faster than nominal wage growth across the bottom three income quintiles. Into this gap, gambling platforms insert themselves as both cause and supposed solution, offering the chance to win what employment cannot provide. The structural similarity to Brazil’s situation is not exact; South Africa’s formal online gambling market is smaller relative to GDP, and its social safety net operates differently. But the underlying pressure is the same: easy digital access to wagering products, aggressive marketing that obscures expected value, and a population segment for whom any loss is catastrophic.
Franc, a local wealth-building platform, has framed the core issue explicitly. The company argues that the true cost of gambling expansion falls heaviest on those least able to absorb losses, and that financial education represents a more durable intervention than prohibition. This position acknowledges what Lula’s ban does not: that demand for gambling will not disappear when supply is restricted, and that illegal markets typically flourish when legal ones are eliminated.
What a Ban Actually Destroys
Brazil’s experience offers a live experiment in the costs of sudden prohibition. The immediate fiscal damage is calculable. The R$8.7 billion collected in federal tax through July 2026 was tracking toward an annual total of roughly R$15 billion, nearly double the previous year. That revenue vanishes entirely. The R$30 million licence fees, already paid by dozens of operators, are forfeited without compensation. The employment effects, while smaller than in physical casino jurisdictions, include thousands of technology, customer service, and marketing positions that were created specifically to serve the regulated market.
More significant is the market displacement. Brazil’s 25 million licensed users will not stop gambling because their apps stop working. They will migrate to unlicensed platforms operating from Curaçao, Malta, or entirely unregulated jurisdictions. These platforms offer no self-exclusion mechanisms, no deposit limits, no Portuguese-language responsible gambling resources, and no tax contribution to Brazilian public services. The government’s own data suggests the black market already handled substantial volume before regulation. That volume will now become the entire market.
Lula’s calculation appears to be that the social cost of regulated gambling, R$38.8 billion by official estimate, exceeded both the tax benefit and the enforcement challenge of controlling black market growth. This is a defensible political judgment. It is not obviously correct as policy. The R$38.8 billion figure bundles together crime, family breakdown, productivity loss, and health expenditure in ways that resist precise measurement. The 1.3 million self-exclusion requests, by contrast, are concrete evidence of harm that regulation failed to prevent. Whether that failure justifies abolishing regulation entirely, rather than strengthening it, is the operational question Brazil now faces and South Africa must consider before reaching a similar impasse.
Alternatives to the Nuclear Option
South Africa has time Brazil did not use. The provisional measure mechanism allowed Lula to act without legislative debate, and the six-day withdrawal deadline reflected that urgency. A parliamentary system with fuller deliberation might have produced transitional arrangements, operator consultations, or phased implementation. It might also have produced nothing, as lobbying and legislative gridlock preserved an unacceptable status quo. Brazil’s choice was constrained by available tools, and those tools produced maximum disruption.
For South Africa, the policy toolkit remains broader. Financial education represents the most frequently advocated alternative, and also the most difficult to implement effectively. School curricula already struggle to cover basic numeracy and compound interest. Adding gambling probability and behavioural economics would require teacher training, material development, and assessment integration that takes years to deploy. Adult financial literacy campaigns, meanwhile, compete for attention in a media environment saturated with precisely the gambling advertising they would counter.
The more immediate lever is advertising control. The Advertising Regulatory Board’s Code of Advertising Practice already contains provisions against misleading claims and appeals to children. Enforcement against digital platforms, particularly social media influencers and programmatic advertising, has lagged. Specific measures that could strengthen the framework without banning gambling outright include: prohibiting advertisements during live sports broadcasts before 21:00, when child viewership peaks; banning celebrity endorsements that imply gambling skill or success; requiring prominent expected-value disclosures in all promotional material; and restricting bonus offers and “risk-free” bet promotions that exploit loss aversion.
Operator-side obligations could also tighten. Mandatory deposit limits, set by default to weekly rather than monthly cycles, would reduce the speed at which losses accumulate. Real-time spending notifications, delivered by SMS rather than in-app messages that users disable, would interrupt dissociative play. Self-exclusion registers, currently fragmented across provincial regulators, could be unified nationally and extended to cover unlicensed operators through payment blocking agreements with domestic banks. The National Gambling Board has explored several of these mechanisms, but implementation has been slow.
The Political Economy of Prevention
Lula’s ban carries electoral timing that cannot be ignored. The first round of Brazil’s general elections falls on October 4, two days before the gambling block takes effect. The provisional measure, valid for 120 days unless Congress acts, positions gambling as a live moral issue during campaign season. Candidates must declare themselves for or against the ban, for or against the Supreme Court challenge, for or against the revenue loss. Lula has forced the conversation on terms favourable to his Workers’ Party coalition, which has historically emphasised state protection of vulnerable populations against market forces.
South African politics offers no direct parallel, but the underlying dynamic translates. Gambling regulation sits at the intersection of provincial and national competence, with the Western Cape and Gauteng maintaining more developed licensing frameworks than other provinces. National legislation would require provincial buy-in that has proven elusive. Meanwhile, the fiscus faces persistent revenue shortfalls, and sin taxes on alcohol and tobacco have approached politically unpopular levels. A regulated gambling expansion, with its promise of new tax streams and job creation, has attracted intermittent interest from Treasury officials. The Brazil example complicates that case by demonstrating how quickly a revenue-positive industry can become politically toxic.
The deeper question is whether any democratic government can sustain a policy that visibly enriches itself from the losses of its poorest citizens. Brazil’s answer, delivered through Lula’s provisional measure, is that the contradiction eventually becomes unsustainable. South Africa’s answer, so far, has been to tolerate the contradiction while expanding regulation slowly enough to avoid confronting it directly. That patience may be running out. The debt statistics, the advertising saturation, and the growing visibility of gambling-related harm in communities already under economic stress are creating pressure that incremental policy adjustment may not relieve.
What Comes After the Ban
Brazil’s operators will not surrender their infrastructure quietly. ANJL’s Supreme Court challenge argues that Provisional Measure 1,394 exceeds constitutional authority, that the six-day withdrawal period violates due process, and that the licence forfeiture constitutes uncompensated expropriation. The court’s timeline is uncertain. If it suspends the measure within the 120-day congressional window, the market could theoretically resume before operators have fully dismantled their operations. If it upholds the ban, the precedent will embolden prohibitionist movements across Latin America and beyond.
For South African policymakers, the litigation outcome matters less than the underlying demonstration. Brazil tried regulated expansion, collected substantial tax revenue, and then reversed course when the social indicators turned negative. The reversal was expensive, chaotic, and legally contested. It was also, by Lula’s framing, morally necessary. The challenge for South Africa is to reach a stable regulatory equilibrium without passing through the same cycle of boom, damage, and sudden closure.
That equilibrium likely requires accepting lower revenue than maximal expansion would produce. Stricter advertising limits reduce operator marketing efficiency and thus market growth. Mandatory player protections increase compliance costs and reduce problem gambler lifetime value, which operators naturally resist. Slower licence issuance and higher entry requirements limit competition and tax base. These are features of a sustainable system, not bugs. The alternative is Brazil’s path: maximal extraction until the political system recoils, followed by maximal destruction.
Franc’s advocacy for financial education and advertising control points toward this middle terrain. The company’s commercial interest, in wealth-building tools for retail investors, aligns with its public position that gambling diversion represents lost compounding opportunity. This alignment does not invalidate the argument. A 25-year-old who directs R500 monthly from sports betting to a low-cost equity index fund will, over a working lifetime, accumulate substantially more wealth than the median bettor extracts from occasional wins. The mathematics are not complicated. Making them culturally resonant, against the constant drumbeat of immediate gratification that gambling marketing provides, is the harder task.
The Harder Task
South Africa’s gambling policy debate has historically been dominated by industry proponents and religious prohibitionists, with limited middle ground. The Brazil example introduces a third position: regulated existence with constrained growth, accepting that some revenue potential must be sacrificed to prevent the market from becoming socially intolerable. This position lacks the clarity of pure advocacy on either side. It requires ongoing regulatory attention, adaptive policy, and willingness to disappoint both operators who want expansion and activists who want abolition.
The specific measures are available. What remains uncertain is whether South African institutions can implement them before reaching a Brazil-style breaking point. The National Gambling Board’s capacity has been questioned in parliamentary oversight hearings. Provincial regulators operate with inconsistent resources and priorities. The Advertising Regulatory Board lacks statutory enforcement power against offshore digital platforms. Financial education, however desirable, competes for curriculum space and teacher attention against foundational literacy and numeracy gaps that are themselves urgent.
Lula’s ban is a warning, not a template. It demonstrates what happens when regulatory failure accumulates faster than political patience. South Africa’s task is to build the regulatory capacity and political will to manage gambling’s growth without requiring such drastic correction. The R$36.9 billion that vanished from Brazil’s economy in days represents not just lost revenue but failed governance. It shows a market that grew faster than the institutions meant to contain it. Whether South Africa can do better will be measured not in headlines but in debt statistics, in self-exclusion registrations, in the slow accumulation of evidence that the industry is paying its way without preying on those who can least afford to lose.
SASFA is an independent coalition, not a statutory regulator. Articles on this blog are commentary and information, not legal advice or an endorsement of any operator.