CNC petition to STF contradicts its own betting figures

The Confederação Nacional do Comércio de Bens, Serviços e Turismo (CNC) filed two petitions with Brazil's Supreme Court on 5 October in defence of MP 1.394/2026, according to a fact-check by BNLData. In ADI 8.024, brought by ANSEJA, the confederation seeks to enter as amicus curiae in a 19-page filing; in ADI 7.721, which it brought itself, it asks for the case to proceed. BNLData said it had access to both petitions and checked their main claims.

The petitions state that spending on betting exceeded R$30bn per month, a figure BNLData rated false and contradicted by the petition's own sources. The report said the figure appears in paragraphs 40 and 70, and in paragraph 76 becomes a flow of about R$30bn per month in bets, which the report described as the correct term: gross volume transferred to platforms, including money returned as prizes. Among the CNC's cited sources, item 4 refers to a Comsefaz bulletin showing average net outflows of R$4.7bn per month, and item 11 cites the Banco Central with R$20.8bn in gross transfers in August 2024. According to official Finance Ministry data, effective spending (GGR) was R$36.9bn in 2025, or R$3.1bn per month, and R$20.07bn in the first half of 2026, or R$3.3bn per month. BNLData said the CNC multiplies real spending by about ten.

A claim that defaults generated by the platforms withdrew R$143bn from retail was rated unsustainable. BNLData said the calculation covers January 2023 to March 2026, while the CNC itself states in paragraph 59 that the regulated market began in January 2025; of the 39 months, 24 predate the regime the entity blames, and 12 predate Law 14.790 of December 2023. The report also cited an internal contradiction: paragraph 14 refers to a loss of R$117bn per year, while R$143bn over 39 months equals R$44bn per year, and in January 2025 the CNC had reported R$103bn for 2024 alone. It added that R$143bn over 39 months equals R$3.7bn per month, more than the entire GGR of the regulated market, and that in 2024, the year the CNC cites a R$103bn loss, retail grew 4.7% according to the IBGE.

BNLData rated as misleading the claim that data from various sources converge to the same conclusion. Of the 12 items in chapter VII, three are the same CNC study (items 1, 2 and 3) and nine are press reports, while the two Banco Central figures (items 10 and 11) are from August 2024, before the regulated market. The report said the health figures do not converge either: growth in SUS attendances appears as a 140% rise between 2018 and 2025 in item 5, a 140% rise since legalisation in item 8, and a rise of more than 600% in item 7.

The report also examined the petition's statement that the Court had unanimously recognised insufficient protective measures. BNLData rated this misleading, noting that the phrase appears in paragraphs 37, 72, 74 and 85, while the petition itself says in paragraphs 28 and 65 that the Supreme Court recognised the fumus boni iuris, the plausibility of the allegation; a provisional ruling is a provisional judgment, the Plenary kept Law 14.790/2023 in force, and the merits were never judged. On the claim that the grant remunerated the period of exploration actually enjoyed, rated false, BNLData said the grant under Law 14.790/2023 was paid for five years of authorisation, while the regulated market operated from January 2025 to October 2026, about 21 months, leaving more than three years paid for but not used. The report said the argument in paragraph 59 that there is no legitimate expectation because the law had been challenged since 2024 ignores that the Executive itself granted the authorisations and received the grants after the action was filed.

On the claim that less severe measures were tested and were insufficient, BNLData found no evidence: the petition, in paragraph 64, presents no assessment of the provisional measures or of the regulation, basing itself on the MP's explanatory memorandum and a 2024 technical note that predates the start of the regulated market. The report said the reading of Nota Técnica 3660/2024/MF is also extended: according to the petition, the note addressed the difficulty of preventing the use of funds from social programmes in betting, and the CNC concludes that the State assumed the structural insufficiency of the authorisation model. On the claim that the MP's mechanisms prevent migration to the clandestine market, rated unsustainable and contradicted by the petition, BNLData noted that paragraph 63 treats the question as settled while paragraph 81 lists the effective capacity of the MP's mechanisms to prevent migration among the fact questions of high complexity requiring a public hearing; site and payment blocking already existed under the regulated regime, and the illegal market kept more than 40% share.

On the claim that a prior public hearing causes no harm to the applicant, rated false, BNLData said the CNC justifies it in paragraph 84 on the basis that the MP already produces its effects, which the report described as precisely the harm: companies are closed and each week without a decision consolidates the prohibition. The entity asks for a new hearing while also seeking to use the November 2024 hearing, which it says has already given the Supreme Court an extensive technical record. On the argument that ANSEJA lacks standing, BNLData described it as a legal argument with an omission: the nine-state rule exists (ADI 4.384-AgR) and the CNC says ANSEJA declares associates in four states, while the exception in ADI 2.866-MC is debatable and will be decided by the rapporteur. The report said paragraph 26 errs in stating that the MP is under discussion in ADIs 7.721, 7.723 and 7.749, since those actions challenge Law 14.790/2023 rather than the MP, and the petition omits ADIs 8.027 and 8.028, also against the MP and also assigned to Fux.

BNLData rated as true the petition's thesis in ADI 7.721 that an MP does not revoke a law but only suspends its effect, following Supreme Court case law (ADI 5.709). The report said this applies to both sides: if the MP lapses or is suspended, Law 14.790/2023 returns in full. In its conclusion, BNLData said the CNC is correct on one point of procedural law and wrong on the facts: the central figure of R$30bn per month is contradicted by two sources the confederation itself attached, the retail loss adds two years in which the regime under attack did not exist, and the hypothesis of the April study reached the Supreme Court as fact.