AI and Consumer Protection: Safeguards Against Abusive Practices

This page is a translation of the original Portuguese text, provided for convenience. It deals with Brazilian law. If there is any discrepancy, the Portuguese version prevails. Read the original in Portuguese

Artificial intelligence already influences purchasing, credit, and customer service decisions, and every consumer should know the mechanisms that protect against abusive practices.

The Rise of Artificial Intelligence in Consumer Relations

We are living in a time when artificial intelligence systems touch nearly every stage of the relationship between businesses and consumers. From personalized offers on digital platforms to automated credit analysis, customer service chatbots, and dynamic pricing, algorithms are an increasingly constant presence in the daily lives of Brazilian consumers. This shift brings undeniable benefits, such as faster service and greater personalization, but it also introduces risks that the legal system must confront seriously and effectively.

Looking at the current landscape, many consumers are not even aware that their commercial interactions are mediated by automated systems. The opacity of algorithms creates an information asymmetry even deeper than the one that has traditionally marked consumer relations. The supplier, who already held a technical and economic advantage, now has tools capable of processing massive volumes of personal data to steer behavior, segment audiences, and, in more serious cases, carry out practices that the Consumer Protection Code classifies as abusive.

In this context, protecting consumers from AI systems does not necessarily require an entirely new legal framework. The CDC (Brazil’s Consumer Protection Code), enacted in 1990, established principles broad enough to cover new technologies: objective good faith, transparency, a ban on abusive practices, and strict liability of the supplier. The challenge lies in interpreting and applying those principles in light of the particular features of algorithmic technology, so that innovation does not become a tool for violating rights.

Abusive Practices Amplified by Artificial Intelligence

The abusive practices listed in Article 39 of the Consumer Protection Code take on new dimensions when carried out through artificial intelligence systems. Dynamic (or personalized) pricing is one of the most debated examples. Algorithms can adjust prices in real time based on the consumer’s profile, considering variables such as geographic location, browsing history, the device used, and even the estimated urgency of the purchase. When this results in unjustified price differences for the same product or service, the situation may qualify as abusive discrimination.

Another troubling phenomenon is the algorithmic “dark pattern,” in which digital interfaces are designed with the help of AI to push consumers into decisions they would not make freely and with full information. Examples include automatically adding products to a shopping cart, deliberately making it hard to cancel subscriptions, displaying false signs of scarcity (“only 2 left”), and manipulating the order in which options appear to steer choices toward those that are more profitable for the supplier. These practices directly violate the principle of objective good faith and the right to adequate information.

In credit and insurance, the risks are just as significant. Automated scoring systems can reproduce and amplify discriminatory biases found in the historical data used to train them. As a result, a consumer may be denied credit or offered unfavorable terms not because of their actual financial capacity, but because they belong to a demographic group that the algorithm statistically associates with higher default rates. This form of algorithmic discrimination, often invisible even to the consumer, violates not only the CDC, but also the Federal Constitution and the General Data Protection Law.

Protecting consumers against abusive AI practices does not require reinventing the law. It requires firmly applying the established principles of good faith, transparency, and strict liability to the algorithmic world.

Automated customer service systems also deserve attention. When companies replace human support entirely with chatbots that cannot handle complex requests, they put artificial barriers in the way of exercising rights. A consumer who needs to file a complaint, cancel a service, or exercise the right of withdrawal may get trapped in automated loops that lead to no real solution, amounting to an abusive practice by obstructing the exercise of rights.

The Legal Framework for Artificial Intelligence and Consumer Protection

Brazil is working toward a specific regulatory framework for artificial intelligence, with bills pending in the National Congress that seek to establish principles, rights, and obligations for the development and use of AI systems. Whatever the legislative stage, there are already legal instruments that apply to consumer protection in algorithmic contexts. The CDC remains the backbone of that protection, complemented by the LGPD (Brazil’s General Data Protection Law, Law No. 13.709/2018), which contains provisions directly relevant to the topic.

Article 20 of the LGPD deserves special attention because it gives data subjects the right to request a review of decisions made solely through automated processing of personal data that affect their interests. This includes decisions about personal, professional, consumer, or credit profiles, or aspects of personality. When we combine this provision with the rights set out in the CDC, we get a robust system of protection: consumers have the right to know that an algorithm made a decision, to understand the criteria used, and to challenge the outcome.

The civil liability of a supplier that uses AI in consumer relations follows the strict liability regime set out in Articles 12 and 14 of the CDC. This means the supplier is responsible for the harm its automated systems cause, regardless of fault. If a pricing algorithm sets an abusive price, a scoring system denies credit in a discriminatory way, or a chatbot blocks the exercise of a right, the supplier is liable for the resulting harm, even if it never intentionally programmed that behavior. The technical complexity of the system is not a defense.

We believe that algorithmic transparency is a fundamental pillar of this discussion. Consumers have the right to be told, clearly and accessibly, when they are interacting with an automated system, which personal data feeds its decisions, and which general criteria guide those decisions. This transparency does not necessarily mean disclosing the algorithm’s source code (which could conflict with trade secrets), but it does require the supplier to be able to explain the system’s general logic in understandable language.

Consumer Defense Mechanisms in the Algorithmic Age

When AI systems carry out abusive practices, Brazilian consumers have several defense mechanisms available. The first and most accessible is filing a complaint with consumer protection agencies, such as state and municipal Procons, and on the Consumidor.gov.br platform. These channels allow mediation between consumer and supplier and, by consolidating complaints, help identify patterns of abusive conduct that may lead to class actions or administrative proceedings.

In the courts, individual and collective actions remain essential tools. The Public Prosecutor’s Office (Ministério Público), the Public Defender’s Office (Defensoria Pública), and consumer advocacy associations can bring public civil actions to stop abusive practices that affect groups of consumers. We consider this route particularly important in the AI context, because algorithmic practices tend to reach large numbers of people at once, making collective protection more efficient than individual claims.

The reversal of the burden of proof, provided for in Article 6, item VIII, of the CDC, takes on even greater importance in disputes involving AI. When a consumer claims to have been the victim of algorithmic discrimination or of an automated abusive practice, it would be unreasonable to require them to technically demonstrate how the system works internally. It is up to the supplier to prove that its algorithm operates within legal parameters, does not discriminate without justification, and respects consumer rights. This allocation of the burden of proof is consistent with the consumer’s informational vulnerability in the face of opaque systems.

We also recommend that consumers take proactive steps to protect themselves: carefully reading privacy policies (especially the sections on automated decisions), regularly exercising their LGPD rights (access, correction, and deletion of data), documenting their interactions with automated systems through screenshots and records, and comparing prices and terms across different devices and profiles to identify possible discrimination.

The Outlook for AI Regulation in Consumer Relations

The regulatory debate over artificial intelligence is moving forward worldwide, and Brazil is following that trend with growing attention to the consumer dimension. The European Union’s experience with the AI Act (the Artificial Intelligence Regulation) offers important reference points, especially its classification of AI systems by risk level and its imposition of obligations proportional to that risk. Systems that directly affect consumer relations, such as credit scoring and personalized pricing, tend to be classified as high-risk and subject to stricter requirements for transparency, human oversight, and impact assessment.

In Brazil, we see that the National Consumer Secretariat (Senacon) and the National Data Protection Authority (ANPD) play complementary roles in overseeing abusive algorithmic practices. These agencies are likely to coordinate more closely in the coming years, issuing specific guidelines on the use of AI in consumer relations, conducting algorithmic audits, and imposing sanctions on suppliers that violate consumer rights through automated systems.

We consider it essential that AI regulation in Brazil strike a balance between innovation and protection. Algorithmic technology can benefit consumers when it is used ethically and transparently, offering services that are more personalized, efficient, and accessible. The goal is not to restrict the use of AI itself, but to stop it from being used for practices that harm consumers, undermine their freedom to make their own decisions, or perpetuate unjustifiable discrimination. Effective regulation should give companies incentives to develop and use AI responsibly, while also ensuring effective means of redress when rights are violated.

As informações deste artigo são de caráter informativo e não substituem consulta jurídica individualizada. Cada caso possui particularidades que exigem análise específica.