Algorithmic Law in Contemporary Brazil: From Cyber-Libertarianism to Constitutional Rights

Review

Pedro Rubim Borges Fortes traces Brazil's shift from a cyber-libertarian conception of online freedom as freedom from state intervention toward digital constitutionalism, in which legislation, regulators, and courts have affirmative duties to protect fundamental rights against algorithmically mediated harm. The article surveys the 2014 Marco Civil da Internet, the 2018 General Data Protection Law, the proposed AI statute, electoral-justice interventions against disinformation, and the 2025 child-protection statute. Its central case is the Brazilian Supreme Court's June 2025 revision of intermediary liability under Article 19 of the Marco Civil. The material is doctrinal and interpretive rather than an empirical assessment of whether the new arrangements produce safer or freer online participation.

The Supreme Court case makes the institutional stakes concrete. Rejecting Article 19's general requirement of a prior judicial order as insufficiently protective, the Court developed a differentiated regime. After extrajudicial notice, platforms can incur liability for failure to remove certain unlawful material, while reputational offenses retain judicial-order protection. Paid amplification and coordinated unlawful bot distribution create distinct presumptions, and specified serious crimes trigger proactive duties directed at systemic failure. The judgment also calls for accessible reporting channels, reasons and appeals for moderation, transparency reporting, and a Brazilian legal representative. Fortes notes that the Court positioned its intervention as provisional pending legislation by Congress. This is not a simple replacement of platform immunity with strict liability: it redistributes the burden of detecting, classifying, and responding to unlawful conduct, with differing evidentiary and procedural thresholds.

The article's useful conceptual insight is that privately implemented code already structures the real exercise of rights; leaving that infrastructure beyond public scrutiny is itself a governance choice. The author's account of an 'algorithmic responsibility of the state' follows from this premise and extends beyond disclosure to investigative capacity, auditing of deployed data processing, and legal access to proprietary systems subject to confidentiality. The account also preserves the crucial distinction between a Senate-approved AI bill and binding law. Brazil's digital constitutionalism emerges here as a layered and still evolving system of statutory duties, judicial remedies, administrative powers, and election-specific controls rather than one coherent enactment.

The analytical difficulty is that constitutional authorization does not establish the legitimacy of every intervention made in its name. The paper describes electoral restrictions and accelerated takedown powers principally as responses to threats to democracy; it offers less systematic scrutiny of error rates, over-removal incentives, political selectivity, adequate notice to speakers, independent appeal, or the asymmetry between powerful and ordinary complainants. The Supreme Court's due-process expectations partially answer this concern, but their effectiveness depends on platform implementation and external review. Similarly, compelling access to algorithms and processing facilities requires clear mandates, confidentiality protections, audit competence, evidentiary standards, and remedies against excessive state intrusion. Those institutional conditions are acknowledged more than specified.

The essay was completed in October 2025 and published in 2026, so its treatment of pending legislation and developing judicial practice should be read as a time-bounded legal account, not a continuously updated statement of Brazilian law. Its broad cyber-libertarianism-to-rights narrative is an illuminating interpretive frame, but it compresses competing constitutional visions into a directional transition and does not demonstrate the practical distribution of benefits and burdens from the new rules. The next evidentiary step would compare actual moderation decisions, appeal outcomes, enforcement consistency, and judicial review across categories of content and platform types. The paper establishes why algorithmic infrastructure has become a constitutional governance object; it does not yet establish that the institutions newly empowered to govern it will be sufficiently contestable, proportionate, or independent.

Key Insight

Digital constitutionalism changes who can govern platform conduct, but the legitimacy of expanded protective powers depends on enforceable limits, reviewable moderation decisions, and remedies for both unlawful content and wrongful suppression.