The Conflict between Freedom of Expression and Other Fundamental Rights and the Issue of Fake News: An Approach According to the Brazilian Federal Constitution ()
1. Introduction
Freedom of speech, among the rights described in the Federal Constitution, is one of the cornerstones of a Democratic State of Law and, therefore, is considered a fundamental right essential to human dignity.
Intrinsically linked to this right is the expression of thoughts and opinions through social media and other internet applications. It is known that modernity has brought to individuals and society as a whole a series of conveniences and forms of connectivity with virtually the entire world.
In this context, the internet, with its social networks and applications, has become one of the main instruments for the expression of thoughts and, consequently, the exercise of freedom of expression.
However, despite the positive aspects that this digital market has brought to society, such as ease of communication and access to information, there are also negative aspects, mainly related to the limits of freedom of speech as a fundamental right and individual and social accountability.
Indeed, posts with offensive, degrading, prejudiced, or hate speech against individuals or groups are common. Additionally, there is a proliferation of posts with a clearly deceptive intent or providing false information, known as fake news.
All of this leads social media managers and internet application providers, in general, to establish rules, known as terms or usage policies, that prospective users must accept if they wish to use them. Among these rules are often norms that prohibit such offensive or deceptive content, and they stipulate penalties for users who violate them, ranging from the removal of the post to the suspension or deletion of the account. This is what is called moderation.
It is precisely at this point that Brazil has recently seen the publication of Provisional Measure No. 1068/2021, which aimed to limit the moderation powers of social media managers and application providers. Since this Provisional Measure lost its effectiveness due to formal unconstitutionality, the Federal Executive presented Bill No. 3227/2021, with virtually the same content, that is, to establish rules limiting the power of moderation within social networks.
Both the Provisional Measure and the Bill sparked intense debates in the legal arena regarding freedom of expression and its limits, as well as the possibility of limiting social media moderation and potential interference with the free initiative of the digital market.
In light of this and the significance of this discussion, the research aims to analyze the constitutionality or lack thereof of Bill No. 3227/2021, particularly focusing on the conflict between fundamental rights such as freedom of expression and the rights to privacy, honor, image, and the right to accurate information, as well as the principle of free enterprise.
To achieve this, the adopted methodology will be exploratory-descriptive, aiming to provide a conceptual clarification of interconnected institutions, offering an overall view of a specific fact, especially when dealing with a less-explored topic where formulating precise and operationalizable causal hypotheses becomes more challenging (Gil, 1999) .
Considering the proposed objective and employing a deductive, systemic, and axiological method, the research will first outline a general overview of fundamental rights such as freedom of expression, privacy, honor, and image, as well as the right to accurate information. Subsequently, it will delve into relevant aspects for the study of fake news in the context of social media use.
Furthermore, the research will then bring to light the key aspects of Bill No. 3227/2021, and finally, address the material constitutionality of the Bill through an analysis of the conflict of fundamental rights and the criteria for resolving them, determining which should prevail in the specific case.
2. Fundamental Right to Freedom of Speech, Its Limits, and Consequences: Privacy, Honor, Image, and the Right to Accurate Information
In close connection with the proliferation of false information is the fundamental right to freedom of expression. The Federal Constitution of 1988 provides that Brazilians can express their opinions freely, as stated in Article 5, sections IV, VI, IX, XLII, as well as in Articles 206, II, 215, and 220. But to what extent does this freedom go? Is it an absolute right, or are there limits?
In modern legal systems, freedom of expression, in a broad sense, constitutes a set of rights related to freedom of communication, which, in turn, includes freedom of expression in the strict sense, freedom of creativity and the press, and the right to information, encompassing numerous other aspects and areas.
In this theme, Silva (2014) understands that freedom of communication consists of a set of rights, forms, processes, and means that enable the unimpeded coordination of the creation, expression, and dissemination of thought and information. This is derived from sections IV, V, IX, XII, and XIV of Article 5 combined with Articles 220 to 224 of the Constitution. It encompasses the forms of creation, expression, and manifestation of thought and information, and the organization of media is subject to a special legal regime.
Thus, considering that there are various ways to demonstrate expressions freely, these rights encompass different fundamental freedoms that must be jointly ensured to guarantee freedom of speech in its full sense (de Magalhães, 2008) . For this reason, it can be said that freedom of speech is linked to other rights, such as the right to inform and be informed, the right of reply, cultural freedom, religious freedom, among others. Therefore, the concept of freedom of speech should be as comprehensive as possible, as long as the functioning of the law is guaranteed.
Analyzing the Federal Constitution and its articles, it is evident that the constituent did not impose explicit restrictions on freedom of speech. Therefore, not even the Judiciary can interfere in the exercise of freedom of speech, which may lead to the false conclusion that freedom of speech is complete and absolute.
Regarding this issue, an interesting discussion was established in Direct Action of Unconstitutionality for Fundamental Precept No. 130 concerning Brazilian Federal Law No. 5520/67 (Press Law) and its infringement on freedom of expression. In this judgment, it was established that freedom of the press is an “overright,” and its limitation is possible when it clashes with another fundamental right. However, this potential restriction could not occur beforehand, as state intervention must take place subsequently, through the Judiciary (Brazil, 2009) .
Thus, it was established that freedom of the press is not compatible with a law made with the intention of restricting it or creating difficulties for the exercise of this political institution. Any law aimed at regulating freedom of the press cannot take on a repressive character that completely distorts it (Brazil, 2009) .
In this sense, it was established that, in the context of freedom of expression, prior censorship is entirely prohibited. On the other hand, such a right is not unlimited and absolute, and the author can be held civilly and criminally responsible, especially when violating the dignity of another person, with emphasis on their personality rights.
Therefore, the answer to the posed question lies within the constitutional text itself, making it clear that the right to the expression of thought, and consequently freedom of expression, is not absolute and has limits. Individuals may incur civil and criminal penalties when their speech and text harm the dignity of another person.
It is important to highlight that the Federal Constitution, in its Article 1, III, presents the dignity of the human person as one of the foundations of the Federative Republic of Brazil. Elevated to the status of a principle and foundation of the Brazilian Republic, the dignity of the human person becomes the essence of a Constitutional State and reflects and impacts various areas of legal knowledge (Marchetti Filho & Marchetti, 2017) .
This paradigm shift in the dignity of the human person and the entire framework involving fundamental rights within the 1988 Constitution demonstrates not only a formal change but a spiritual one by the original constituent, placing the human person at the center of concerns rather than the organization of the State, with its competencies and structure (Martins, 2021) .
Therefore, it is correct to say that “in addition to being a principle, as already mentioned, the dignity of the human person is inserted into the legal scene as a fundamental guarantee of the individual, and by this idea, it must apply to all areas of legal knowledge” (Marchetti Filho & Marchetti, 2017) .
Despite the difficulty in conceptualization, given its vague and imprecise contours, coupled with a touch of ambiguity and porosity (Sarlet, 2015) , exacerbated by the inherently polysemic nature of the expression (Marchetti Filho & Marchetti, 2017) , we can say that, for this research, human dignity is linked to the intrinsic and distinctive quality of each human being, protecting them from any “degrading treatment and odious discrimination, as well as ensuring minimum material conditions for survival.” Thus, human dignity is an “attribute that every individual possesses, inherent to their human condition, regardless of any other condition related to nationality, political preference, sexual orientation, creed, etc.” (Ramos, 2017) .
Therefore, as an intrinsic right of the individual, it must be respected and, at the same time, serves as both a sustaining and limiting factor for freedom of expression. Analyzing its context, we can point out that “freedom of expression, despite being broad, is not absolute and unlimited, as the freedom of each person has as its limits the freedom and rights of others” (Araujo, 2018) .
In this vein, considering that fundamental rights have the nature of principles with a value-laden and axiological content, freedom of expression, as a species, is limited within the dignity of the person by personality rights, especially privacy, honor, and image.
Privacy is a manifestation of the right to intimacy and is embodied in the “demand for respect for the isolation of each human being, who does not wish for certain aspects of their life to be known to third parties. In other words, it is the right to be alone” (Gagliano & Pamplona Filho, 2016) . This right is expressed in situations involving the home, family, and correspondence, for example (Marchetti Filho & Pereira, 2021) .
As its more intimate derivative, there is the right to secrecy, understood as that which “we do not want to reveal or make public, that is kept in the deepest recesses of the mind, in the innermost part of the person,” viewed in terms of communications, the home, and the profession (Marchetti Filho & Pereira, 2021) .
On the other hand, the right to honor is “one of the main aspects of personality. This is because honor is intrinsically linked to human nature. It is born with the person and accompanies them until their death. Sometimes even after their death” (Marchetti Filho & Pereira, 2021) , and it can manifest in a subjective manner, “located in the innermost recesses of their affections and feelings,” and objectively, reverberating in “the society in which the person lives and is inserted (objective honor)” (Bentivegna, 2020) .
Linked to honor, we have the right to image, which “constitutes the sensible external expression of human individuality, worthy of legal protection” (Gagliano & Pamplona Filho, 2016) . In fact, “the image corresponds to the externalization of personality, encompassing, at the same time, the physiognomic reproduction of the holder and the sensations, as well as the behavioral characteristics that make them particular, distinguished in social relations” (Rosenvald & de Farias, 2015) .
This broad concept brings the distribution of the image within a trilogy, presented in doctrine as the trilogy of portrait/attribute/voice (Marchetti Filho & Pereira, 2021) , and arises from the constitutional protection of honor and image in Article 5, paragraphs V and X, which states that “the right to reply, proportional to the offense, as well as compensation for material, moral, or image damage, is ensured.” Therefore, “the intimacy, private life, honor, and image of individuals are inviolable, with the right to compensation for material or moral damage resulting from their violation” (Brazil, 1988) . This constitutional protection is regulated and individualized in Article 20 of the Civil Code.
These rights, within the scope of human dignity, limit the right to freedom of expression. But that’s not all. There is also the right to accurate information, which ensures that individuals have access to information with correct content, whether of collective interest or private interest, that effectively aims to inform and not deceive or cause misinformation.
Therefore, the purpose of this right is to make information accessible, providing clarifications on specific subjects, as well as the means to obtain it, but in a correct manner and from reliable and secure sources, except for those that are classified as confidential by law. After all, freedom of expression and information are pillars of a well-structured democracy within the values of citizenship.
Understanding this, when the expression of thought, within freedom of expression, violates an individual or collective right, it results not only in criminal liability if the act constitutes a crime but also in civil liability for the individual or social damages that may arise.
It is known that “civil liability is the legal duty that arises from the violation of another legal duty and obliges the wrongdoer to compensate for the resulting damage. It arises to restore the damage resulting from the violation of an obligation present in the law or in a contract.” Thus, “responsible is the person who must compensate for the damage resulting from the violation of a pre-existing obligation, that is, a pre-existing legal or contractual duty that has been breached” (Marchetti Filho, 2018) .
In this context, civil liability is based on the idea that anyone who harms a legal right through a pure wrongful act (CC, Article 186) or through an abuse in the exercise of a right (Article 187) must compensate for it. In this context, if a particular publication has content that is offensive to the privacy, honor, or image of an individual, or content that is knowingly false and disseminated with the intention of deceiving, misleading, or concealing the truth, and if it causes individual or social harm, the duty of compensation arises, in accordance with Article 927 of the Civil Code.
Consequently, it is correct to affirm that the fundamental right to freedom of expression, present in the constitutional text, is broad and does not allow prior censorship. However, it requires individuals to exercise care because this expression of thought is not absolute and is limited by other fundamental rights related to human dignity and personality, such as privacy, honor, and image, as well as the right to accurate information. If there is individual or social harm due to the publication of offensive content to these rights or containing false information, it can lead to civil liability and the corresponding obligation to compensate, among other consequences.
3. The Issue of Fake News and Social Media as a Propagation Instrument
The expression “fake news” originates from the English language and means “false news.” These are false news pieces in which tactics are employed to give them the appearance of truth. They are generated by mass media and published with the intent to deceive, gain financial or political advantages (Sobral, 2018) .
Therefore, in a general definition, fake news refers to false information disseminated with the intention of persuading people to believe in a certain fact as vehemently true, leading them into error, or shaping a tainted opinion or will.
It is observed that the primary mode of dissemination of fake news is closely tied to modernity: they are “shared on the internet as if they were true, mainly through social media and messaging apps,” usually with the aim of “creating controversy around a situation, garnering visits or views on websites and internet videos” (Flumigan & Lisboa, 2020) .
However, there are still other aspects directly linked to this. Fake news originates from the full freedom of individuals to express themselves freely, and social media has been a fast communication channel, where anyone can have access, spreading various information that may have false content (Gutierrez, 2020) .
In 2016, this term gained recognition and took on a global scale with the political events during the presidential campaign in the United States. Studies indicate that the American Democratic Party was leading in the polls to win the election. However, some fake news regarding the personal life of the Democratic candidate was spread, degrading her image and causing a considerable drop in her votes (Gutierrez, 2020) .
These fake news spread in regions where Democratic and Republican tendencies were not predominant, influencing people who had not yet chosen their candidate. However, this tactic did not achieve much success with voters who already had a formed opinion about their candidate, but still harmed a side that could not reverse the situation.
As mentioned by Mars (2018) , studies conducted by researchers from Princeton, Dartmouth, and Exeter, as published in January, analyzed responses from 2525 Americans in a computer-recorded traffic study. They found that one in four Americans visited fake news during the campaign. However, the consumption of these rumors was concentrated in a very small group: six out of 10 visits to gossip sites reached 10% of people with more conservative reading preferences, meaning those with a small number of doubts about their vote. However, according to experts, their existence likely did not alter the election results.
Mars (2018) also presents another study conducted by economists Hunt Alcott and Matthew Gentzkow, which was based on a survey of American memory. In it, they emphasized that “no more than 8% of people who saw fake news believed in it, which also suggests that they hardly changed the outcome, but the scholars explicitly warned in their report that they do not have conclusions about it.” However, clearly, the news that circulated on American screens, the fake news disseminated at the time had a direct connection to the final result of the campaign.
However, in Brazil, a significant event occurred in 2014, in which through fake news circulated on social media, a 33-year-old woman named Fabiane Maria de Jesus was tied up and brutally beaten by local residents in the region where she lived, ultimately leading to her death two days later. This happened because she was mistaken for an alleged child kidnapper who apparently never existed (Steil, 2021) .
This fake news was shared through social media without verifying the real truth of the matter. The incident brought indignation to the family, which sought to hold the page accountable for authorizing the dissemination of the material in question (Steil, 2021) . However, the request for compensation for moral damages was denied by the Judge of the 3rd Civil Court of the Central Forum of São Paulo, who understood that “there was no court order for the content to be removed, and the defendant is not the police of the platform’s users’ customs but a mere repairer afterward, according to the terms of use and prior notice,” as per Article 19 of Federal Law No. 12,965, of 2014 (Brazil, 2020a) .
Moreover, in Brazil, the use of fake news has also become a practice in elections, with false news about candidates and, currently, about the security of the electoral process, jeopardizing the democratic process of choosing political representatives, similar to what happened in the United States. It is, therefore, a dangerous practice that needs attention, as it can visibly cause irreversible damage.
Despite the expression and the presented concept being modern, gaining great visibility recently, the idea itself of using and spreading lies to achieve certain objectives is not new. In truth, “fake news is a recurring device throughout history to deceive the opinions of the masses and, together with other social factors, legitimize atrocities,” often with clear political motivation, a fact that is no different today (Santana & Silva, 2019) .
Indeed, Santana and Silva (2019) point out in their research that there are reports of the use of fake news since Ancient Rome, through the Middle Ages, with influences even in stories of “witch hunts,” as well as in the persecution of Jews. “In the 20th century, the spread of fake news had an impact on the World Wars, the rise of the Soviet Union, Nazism, and subsequent wars, such as the Vietnam War (1955-1975) and the invasion of Iraq (2003).”
In modern times, what has truly changed is the way in which such false information is propagated. As mentioned earlier, the medium for spreading fake news is the internet, with its websites, social media platforms, and private messaging apps. Indeed, with its arrival, it is not uncommon to receive false news daily through mobile phones, computers, and television. The content, for the most part, is controversial and often tarnishes the image of someone, an institution, or something significant.
Therefore, it is observed that the primary mode of dissemination of fake news is strictly linked to modernity: they are “shared on the internet as if they were true, mainly through social media and messaging apps,” usually with the goal of “creating controversy around a situation, garnering visits or views on websites and videos on the internet” (Flumigan & Lisboa, 2020) .
The spread of these low-credibility issues can even bring profits to their providers, as it is possible to attract advertising with the repercussion of the news, due to hasty sharing without proper verification of its truthfulness. It is still possible for pages to change small details that were not displayed or were posted out of context, causing harm to third parties.
Furthermore, the speed at which this type of false content is spread also causes concern, considering that in a matter of seconds, any type of content can pass through the hands of internet users and their applications. However, the big tech companies, controllers of most apps, took a while to take action in their self-regulation power to prevent the spread of fake news.
Indeed, it is observed that only from 2017 onwards, Facebook, for example, began to take various actions to block suspicious content, automatically sending these publications for analysis, thus preventing them from being spread (Soares, 2018) . However, these measures have not proven to be 100% effective, allowing a significant amount of content to escape control.
But, just like this social network, others have also shown concern about the issue and provided ways to prevent these disclosures. Belonging to the same group, WhatsApp sought to test a feature that alerts the user that the material in question may be a rumor, based on the number of times it has been forwarded by numerous people who also use the service. Following the same path, Google and Snapchat improved their tools and updated their guidelines (Soares, 2018) .
In another application, in a study reported by Pereira (2019) , “126,000 stories collected from the Twitter application, between 2006 and 2017, about the dissemination of true and false news online, it was found that false news was spread more widely and more quickly than news considered true.” And among these disseminated fake news, the one that ranked first had politics as its subject, “followed by terrorism, natural disasters, science, urban legends, and financial information.”
The study in question highlighted that fake news reached a significantly higher audience compared to true information. While authentic news reached approximately 1000 people, fake news managed to reach up to 100,000 people, indicating that a much larger number of people shared falsehood compared to the truth. The spread of misinformation was facilitated by its virality, suggesting that falsehood spread not only through transmission dynamics but also through point-to-point dissemination, characterized by a viral branching process (Pereira, 2019) .
This concern about the spread of news in the media has reached the Legislative Branch, leading the Federal Senate to adopt measures in an attempt to raise awareness among users about responsibility, credibility, and the quality of information when publishing content. This was the central objective of the campaign “Press Freedom, vibrant democracy” (Monteiro, 2021) .
This same concern also led to the processing of Bill No. 2630/2020, which seeks to establish the “Brazilian Law on Freedom, Responsibility, and Transparency on the Internet,” also known as the Fake News Law (Brazil, 2020b) .
The project itself aims to establish “rules related to the transparency of social networks and private messaging services, especially regarding the responsibility of providers for combating disinformation and increasing transparency on the internet” (Brazil, 2020b) .
Thus, it aims to guarantee, as per Article 3, “freedom of expression and the press,” the “rights of personality, dignity, honor, and privacy of the individual,” “respect for the user in their free formation of political preferences and a personal worldview,” and the “reliability and integrity of information systems,” as well as the “promotion of access to knowledge in the conduct of public interest matters” and “broad and universal access to media and information,” along with, of course, “consumer protection” and “transparency in the rules for the dissemination of advertisements and paid content” (Brazil, 2020b) .
In its current stage of proceedings, the project was approved by the Federal Senate and awaits processing in the Chamber of Deputies. However, from it, it is clear that the quest for transparency and cooperation between digital platforms and state agents, along with greater scrutiny of inauthentic behavior in the spread of fake news, is evident.
Always important to remember that “the spread of Fake News typically highlights a commissive conduct. So, the publication of fake news or news containing false elements is a commissive human conduct.” As an exception, one can observe an omissive conduct in a specific case, notably “in cases of journalistic civil liability for failure to verify Fake News when journalistic outlets replicate Fake News from other media” (Guimarães & Silva, 2019) .
Therefore, the dissemination, whether through the creation and initial publication or through the sharing of fake news, constitutes a commissive act, precisely manifested by the action of creating, disseminating, or forwarding the news without verifying its content.
4. The Bill No. 3227/2021, Its Motivation, and Content
The Bill No. 3227/2021 was presented by the Federal Executive after the rejection of Provisional Measure No. 1068, also from 2021, by the Senate and returned by the President of the Senate, Senator Rodrigo Pacheco.
The Provisional Measure aimed to amend Federal Law No. 12,965 of 2014, known as the Brazilian Civil Rights Framework for the Internet, to limit the self-regulation of social networks and service providers in removing content and user accounts without a court order.
For this reason, the Executive’s action was strongly criticized by civil society and the legal community. Firstly, because such a change could never be made by a provisional measure, but rather following the same constructive line as the Marco Civil da Internet, that is, through extensive discussion in a “multi-participatory, multisectoral” manner and based on consensus among different sectors. In other words, any changes to this law should “follow its democratic and representative character. The modification of the Marco Civil through a unilateral act like a Provisional Measure constitutes a setback in relation to the democratic victories that the regulation represents” (Iris, 2021) .
Furthermore, it is certain that the use of the internet to express freedom of thought, “even if intermediated by private services, plays a public space function. Therefore, content moderation is a topic that involves fundamental rights, such as freedom of expression, access to information, and privacy.” For this reason, there is a consensus that there is a “need for greater transparency on how speech is controlled on the internet.” However, establishing a “taxative list of hypotheses for moderation limits the activity of platforms inappropriately, considering the diversity of content, including illicit and harmful content, that circulates and needs to be curbed—such as misinformation.” This is because it is practically impossible to predict all forms of content that require platforms to moderate. Thus, “materially limiting these cases can expose people to harmful content, creating a potentially hostile environment that may even inhibit their expression” (Iris, 2021) .
In conclusion, because the Measure presents a notorious political interest, given that the majority of accounts and publications removed had content related to political agents of the Federal Executive, with the offensive intention towards opponents or the dissemination of false information. All under “an alleged defense of freedom of expression to prevent the fight against disinformation on social networks” (Pereira Junior & Vieira, 2021) .
Its validity lasted about a week when it ceased to be effective due to almost simultaneous decisions by the President of the Federal Senate Rodrigo Pacheco and Supreme Federal Court Justice Rosa Weber, given its blatant formal unconstitutionality (Pereira Junior & Vieira, 2021) .
Therefore, the Executive Branch presented Bill No. 3227 in 2021, with the same content and objective: to amend the Marco Civil da Internet Law, strictly limiting the possibilities of self-regulation by application providers and social networks in removing profiles and content from social networks that manifestly violate their terms of service unless “just cause” is proven.
In its motivation, the proponent emphasizes the justification for the importance of the project:
This absence of specific rules applicable to social networks is even more serious when considering that, in 2021, about 150 million Brazilians use social networks, which corresponds to more than 70% of the population. Faced with this new reality, where social networks play a fundamental role in mediating personal and professional relationships for a majority of the Brazilian population, it is necessary to establish clear rules on the use of these platforms in order to protect users from arbitrary, unilateral, subjective, and secret decisions by social network providers. In particular, it is necessary to prevent these decisions from being made without proper legal process, especially the possibility of opposing views or appealing to the platform itself or the Judiciary.
It is, therefore, the responsibility of the public authorities to ensure compliance with the Constitution and national legislation, especially regarding principles, guarantees, rights, and duties for the use of the internet in Brazil, ensuring that the relationships between users and social network providers occur in a context marked by legal certainty and respect for fundamental rights (Brazil, 2020a) .
However, the reality of the text reveals other clear objectives. Indeed, with content practically identical to the Provisional Measure, one of the points of the amendment lies in the new concepts introduced in Article 5 of the Marco Civil da Internet Law, which, if approved, will be in effect with the following subsections:
Art. 5 For the purposes of this Law, it is considered: […] VII—internet applications—the set of functionalities that can be accessed through a terminal connected to the internet; VIII—records of access to internet applications—the set of information regarding the date and time of use of a specific internet application from a certain IP address; IX—social network—an internet application whose main purpose is the sharing and dissemination, by users, of opinions and information, conveyed through texts or files of images, sounds, or audiovisuals, on a single platform, through connected or articulated accounts, allowing the connection between users, and provided by a legal entity engaged in economic activities in an organized manner, by offering services to the Brazilian public with a minimum of ten million registered users in the country; and X—moderation on social networks—actions by social network providers involving the exclusion, suspension, or blocking of the dissemination of user-generated content and actions of cancellation or suspension, total or partial, of social network user account or profile services and functionalities. The definition in item IX of the caput does not include internet applications intended for instant messaging and voice calls, as well as those whose main purpose is to facilitate the trade of goods or services (Brazil, 2021a) .
Within this new conceptual framework, the Bill adds to the Law the art. 8-A, under the title “rights and guarantees of social network users.” One of the most important points is in items V and VI, and the sole paragraph, as follows:
Art. 8-A Users, in their relations with social network providers, are assured of the following rights, without prejudice to the provisions of Section I of this Chapter: […] V—no exclusion, cancellation, or suspension, total or partial, of services and functionalities of the account or profile, except for just cause, subject to the provisions of art. 8-B; VI—no exclusion, suspension, or blocking of the dissemination of user-generated content, except for just cause, subject to the provisions of art. 8-C. […] Sole paragraph. Social network providers are prohibited from adopting moderation criteria or limiting the reach of content dissemination that implies censorship of a political, ideological, scientific, artistic, or religious nature, subject to the provisions of art. 8-B and art. 8-C. (Brazil, 2021a)
The simple reading of the clauses makes the real objective of the Project clear: to make it difficult to delete content and cancel accounts due to user posts. Analyzing the historical context preceding both the Provisional Measure and the Bill, it is evident that their real motivation was the various deletions of posts and accounts, many of them fake, with offensive content or a high load of fake news, some even involving the President of the Republic and his support network.
This purpose becomes even more evident when we see, in articles 8˚-B and 8˚-C, the specific nature of the circumstances that would constitute just causes for the exclusion of accounts or content posted by users:
Art. 8˚-B In observance of freedom of expression, communication, and the expression of thought, the exclusion, cancellation, or suspension, total or partial, of services and functionalities of the user’s account or profile on social networks can only be carried out with just cause and motivation. § 1 The just cause is considered characterized in the following cases: I—default by the user; II—accounts created with the purpose of assuming or simulating the identity of third parties to deceive the public, except for the right to use a social name and pseudonym and explicit humorous or parodic intention; III—accounts predominantly managed by any computer program or technology to simulate or replace human activities in content distribution on providers; IV—repeated practice of the conducts provided for in art. 8˚-C; V—accounts that offer products or services that violate patent, trademark, copyright, or other intellectual property rights; or VI—compliance with a judicial determination. […] Art. 8˚-C In observance of freedom of expression, communication, and the expression of thought, the exclusion, suspension, or blocking of the disclosure of content generated by the user can only be carried out with just cause and motivation. § 1 The just cause is considered characterized in the following cases: I—when the content published by the user is inconsistent with the provisions of Law No. 8069, of July 13, 1990; II—when disclosure or reproduction constitutes: a) nudity or explicit or implicit representations of sexual acts; b) practice, support, promotion, or incitement to crimes against life, pedophilia, terrorism, trafficking, or any other criminal offenses subject to unconditional public criminal action; c) support, recruitment, promotion, or assistance to criminal or terrorist organizations or their acts; d) practice, support, promotion, or incitement to acts of threat or violence, including for reasons of discrimination or prejudice based on race, color, sex, ethnicity, religion, or sexual orientation; e) promotion, teaching, encouragement, or apology for the manufacture or consumption, explicit or implicit, of illicit drugs; f) practice, support, promotion, or incitement to acts of violence against animals; g) use or teaching of the use of computers or information technology with the aim of stealing credentials, invading systems, compromising personal data, or causing harm to third parties; h) practice, support, promotion, or incitement to acts against public safety, national defense, or State security; i) use or teaching of the use of internet applications, websites, or information technology with the aim of violating patent, trademark, copyright, or other intellectual property rights; j) violation of the rules issued by the National Council of Self-Regulation Advertising regarding advertising or promotional content; or k) dissemination of software viruses or any other computer code, file, or program designed to interrupt, destroy, or limit the functionality of any computer resource; or l) marketing of products unsuitable for consumption, as provided for in §6 of art. 18 of Law No. 8078, of September 11, 1990; III—request from the offended party, their legal representative, or their heirs, in the event of a violation of intimacy, privacy, image, honor, protection of their personal data, or intellectual property; or IV—compliance with a judicial determination (Brazil, 2020a) .
Indeed, from a simple reading, it is clear that all these motivations already constitute grounds for the exclusion of accounts and posts by social network moderators. However, it is evident that, in these cases, there is no provision for the exclusion of accounts or posts that have notoriously false content with the intent to deceive or misinform, let alone those that propagate insults and hatred against political opponents, as well as gender, race, or religious prejudice.
Furthermore, in article 28-A, there is a provision for very heavy sanctions, even reaching the point of suspending or interrupting the activity of the social network that violates the Law, thereby harming the access of all other users of the application.
Finally, just like the Provisional Measure, the Project itself aims to interfere with the usage rules established by the managers of the applications and their moderation power, by analyzing the compliance with these rules, interfering, in addition to other fundamental rights, with the free initiative and self-regulation within the digital market, as we will see next.
5. The Constitutionality of Bill No. 3227/2021: An Analysis from the Perspective of the Conflict between Fundamental Rights
It is certain that many fundamental rights explicitly or implicitly provided for in the Constitution have a principled value load, meaning they are elevated to the category of fundamental principles.
Furthermore, in some situations, conflicts may arise between fundamental rights, with equally relevant principled content, and thus it is necessary to analyze and decide which should prevail in the analysis of the specific case.
In fact, it is common for clashes of ideas or interests that manifest fundamental rights, and as a result, a divergence arises in which the decision between the predominance of one or another fundamental right becomes of utmost importance.
In light of this, according to Alexy (2017: p. 93) , when there are collisions between equally valid constitutional fundamental rights, the solution should be based on the following idea: If two principles collide—which occurs, for example, when something is prohibited according to one principle and permitted according to another—one of the principles will have to give way. However, this does not mean that the yielding principle should be declared invalid, nor that an exception clause should be introduced in it. In fact, what happens is that one of the principles takes precedence over the other under certain conditions. Under other conditions, the question of precedence can be resolved in the opposite way.
“This means that no fundamental right is unlimited. ‘After all, everything has its limits in human affairs.’ Likewise, ‘fundamental rights are not absolute and have limits, especially when the exercise of one right interferes with the right of another person’ (Marchetti Filho & Pereira, 2021) . Therefore, the fundamental rights provided for in the Federal Constitution “are not unlimited, as they find their limits in other rights equally enshrined in the Constitution (Principle of relativity or coexistence of public freedoms)” (de Moraes, 2022) .
That is why, in the specific case, “principles have different weights, and principles with greater weight take precedence”. It should be noted that the method of resolution here is different from rule conflicts. This is because “conflicts between rules occur in the dimension of validity, while collisions between principles—since only valid principles can collide—occur, beyond this dimension, in the dimension of weight” (Alexy, 2017) .
In this line of reasoning, given the evaluative weights between the principles, we understand that “the limits of fundamental rights, when not directly provided for in the Constitution, are demarcated in the abstract by the legislator or in a concrete manner by the constitutional judge” (Barroso, 2015: p. 371) . Therefore, “the resolution of the conflict, in general, must occur specifically in the concrete case by the judge, who must employ proportionality and the weighing of goods and values” (Marchetti Filho & Pereira, 2021) .
Thus, in this solution, the analysis must be carried out, in light of the specific case, through proportionality and “also, in the sense of allowing the judge to assess the weight of the norm in a particular incidence, in order to prevent it from producing an undesirable result for the system, thus achieving justice in the specific case” (Barroso, 2015) .
Therefore, “there cannot be a resolution of the conflict between two principles by suppressing one in the face of the other. The weight and relative importance of each should be considered in the agenda within the specific case, in order to minimize the constraint of one over the other.” In this weighing, “judges and courts should decide the conflict through the weighing of goods and values, which consequently encompasses the principle of proportionality at its core, so that the decision made is the least harmful to the effectiveness of fundamental rights” (Marchetti Filho & Pereira, 2021) .
In this regard, based on Alexy’s premise, Barroso (2015) presents the three phases of the technique for resolving conflicts: In the first stage, it is the interpreter’s task to identify the relevant norms within the system for resolving the case, identifying any potential conflicts between them. In the second stage, it is necessary to examine the facts, the specific circumstances of the case, and their interaction with the normative elements. Thus, the examination of the facts and their impact on the identified norms in the first stage can provide greater clarity on the role of each norm and the extent of their influence. It is in the third stage that the process of weighing becomes distinct, as opposed to subsumption. In this decision-oriented phase, the different groups of norms and the impact of the specific case’s facts will be examined together to determine the weights that should be assigned to the various elements in dispute and, therefore, the group of norms that should be considered in the case. Then, it will be necessary to further decide how intensely this group of norms—and the solution indicated by it—should prevail over the others.
Considering this, we have that the criterion of proportionality with a weighing of values shows that the imposed restrictions must be appropriate to ensure the overriding right, as well as being the least burdensome path for the overlapped right, and that in the end, the desired goal has been achieved (Marchetti Filho & Pereira, 2021) .
Therefore, in the specific context of the research topic, the discussion of Bill No. 3227/2021 brings to light the conflict between the fundamental right to freedom of expression on one side, and the rights to privacy, honor, image, and accurate information on the other, all with the same weight and importance in the constitutional framework.
Therefore, it must be considered that everyone is guaranteed the fundamental right to freedom of expression and, as such, everyone is free to express their thoughts and have their own conscience, and belief.
However, this freedom has limits within the constitutional text itself, either in terms of prohibiting certain actions or in terms of holding individuals accountable for the individual and social harm caused.
It means that the fundamental right to freedom of expression reflects the essence of democracy in terms of the possibility of freely expressing one’s thoughts, as long as it does not violate the fundamental rights of others (Marchetti Filho & Pereira, 2021) . One of the ways this can happen is through the publication of offensive content that infringes upon the privacy, honor, and image of individuals, or through the dissemination of notorious false information, such as fake news. Therefore, these can be considered as limitations to freedom of expression.
In light of this, depending on the specific case, in the evaluative analysis of values within the criterion of proportionality, the prevalence of the fundamental right to privacy, honor, image, and accurate information over the fundamental right to freedom of expression becomes clear when this freedom of expression is used abusively to spread offensive or false news and information with the clear intention of tarnishing the privacy, honor, and image of individuals, or deceiving, changing or forming erroneous opinions, or even creating a false reality about a fact that does not actually exist.
In this circumstance, we have the context of Provisional Measure No. 1068/2021, which was preliminarily rejected by the Federal Senate and had its effectiveness suspended by the Supreme Federal Court due to blatant unconstitutionality. However, this rejection was based on formal grounds and not on the analysis of its content itself.
Indeed, according to article 62 of the Constitution, a Provisional Measure can only be issued by the President of the Republic in cases of relevance and urgency. However, as pointed out by Minister Rosa Weber, the explanatory statement of the referenced Provisional Measure did not demonstrate “in a reasoned and sufficient manner the presence of the urgency requirement, especially in a matter of such complexity and vicissitudes, which evidences the absence of such constitutional requirement, resulting in apparent formal unconstitutionality” (Brazil, 2021b) .
Now, with the presentation of Bill No. 3227/2021, following the rejection of the Provisional Measure on formal grounds, the material constitutionality of the content of this bill needs to be analyzed, notably regarding the conflict between the fundamental rights of freedom of expression, privacy, honor, image, and accurate information.
In these terms, it is understood that the analysis of the content of the Provisional Measure seeks to guarantee freedom of expression in an unlimited manner, which is effectively a mistake. As we mentioned, no fundamental right is unlimited, and freedom of expression is also not exempt from limits imposed by the Constitution itself.
However, by the wording of the bill, especially regarding the issue of content removal and profile deletion by social media platforms and application providers, there is an extreme limitation on the power of action of these managing companies. This is notable because such moderation can only occur when the incident falls within the specific cases of “just cause” listed in paragraphs 1 of articles 8-B and 8-C. This, evidently, practically exposes other users to offenses, hate speech, and, especially, fake news.
Furthermore, the Project’s idea, as it stands, aims to interfere with the management of the rules established by the administrators of the applications when they defined the terms and policies of use, which were accepted by the users when they opened their accounts. Therefore, users are aware, albeit implicitly, of the rules of use.
Truly, by establishing “terms of use and community guidelines,” social networks set the “rules of the game” and aim to differentiate their products to attract users and advertisers, just as any company seeks to improve/differentiate its product to increase sales (Pereira Junior & Vieira, 2021) . If the rules created by a social network result, for example, in hostile environments or marked by discrimination against individuals, there is a significant risk that users and advertisers will migrate to competing platforms. In this sense, the ability to create rules that define the design characteristics of the product offered to users and advertisers is a central element of free enterprise and free competition for social networks (Pereira Junior & Vieira, 2021) .
Now, if the program is managed by its owner and therefore has the power to define rules for its use and manage compliance with them, any initiative to establish regulations that prohibit application administrators from deleting content or accounts in moderation for non-compliance with their rules, in our view, also interferes with free enterprise, a pillar of democracy.
In other words, “by establishing an exhaustive list of situations that constitute ‘just cause’ for the removal of users and content by social networks,” the Project also has a basic profile of direct interference in the design of social networks, “limiting not only their capacity for differentiation (thus reducing a key element of the competitive process) but also the option for legitimate design choices that seem to align with the preference of a large portion of users” (Pereira Junior & Vieira, 2021) .
In this line of reasoning, this interference seeks to replace “the exercise of free enterprise, competition between different designs, and innovation with a centralized decision by the State.” Therefore, it violates another fundamental guarantee that underpins the power of self-regulation of social networks and other applications through content moderation, which does not seem “justifiable due to market failures traditionally considered—market power, externalities, and informational asymmetry” (Pereira Junior & Vieira, 2021) .
All of this, coupled with the fact that this moderation aims to specifically remove posts and accounts with clearly offensive content, disseminators of hate speech and prejudice, or fake news, shows that the content of the referenced Bill, as it stands, is unconstitutional for privileging unlimited freedom of expression, to the clear detriment of fundamental rights such as privacy, honor, image, accurate information, and free enterprise.
Of course, we are not advocating for a complete lack of regulation by the Law, as that is necessary. However, such regulation must respect not only freedom of expression but also ensure that this freedom of expression does not harm other fundamental rights of equal importance, and that offensive or false content should yield to these other fundamental rights. Public debate on topics such as freedom of expression and the constitutional limits of provisional measures is crucial. However, a less emphasized aspect in the discussion that followed MP 1068 is its unconstitutionality for limiting free enterprise and the constitutional principles of free competition and contractual freedom. To address these issues, it is necessary to understand the business model of social networks (Pereira Junior & Vieira, 2021) .
The truth is, behind its lines, this Project has a clear political and opportunistic interest, which, in itself, should completely dismiss the possibility of even considering the project’s progress.
The debate about application providers, social networks, and other internet applications, their rules, and the moderation of their content through self-regulation, is necessary. However, it should be done with caution and in a comprehensive manner, involving the entire community, discussing, among other aspects, “transparent and ethical use of the powers exercised by platforms” (Leal & Palhares, 2021) .
All of this while always observing the fundamental rights involved, in order to avoid any threat or undue interference in the evolution of the dynamics and innovation of social networks and digital markets in general, which, despite some negative aspects, generate numerous benefits to consumers due to their importance in modern social life.
6. Conclusion
The Brazilian Federal Constitution, as a pillar of the Brazilian Democratic State, guarantees a series of fundamental rights to all, including freedom of expression, which must be exercised within the limits established by the constitutional text itself, with responsibility and the possibility of being held accountable in case of abuse in its exercise.
In addition to freedom of expression, the Constitution also guarantees other fundamental rights of equal importance, such as privacy, honor, and the right to accurate information, individual rights that must be respected.
When faced with a conflict between fundamental rights of equal importance, it is necessary to analyze, within a criterion of proportionality and weighing of values, which one should prevail over the other.
In this case, in a conflict between the fundamental right to freedom of expression and the right to privacy, honor, image, accurate information, and free initiative, it seems that, based on the research conducted here, the rights to privacy, honor, image, accurate information, and free initiative should prevail. This is due to the limitations that the Federal Constitution has imposed on freedom of expression.
When analyzing the content of Bill No. 3227 of 2021, in an attempt to reissue the provisions of Provisional Measure No. 1068 of 2021, it is clear that it is based on the false premise that freedom of expression can be exercised without limits.
Within this context, it is understood that, beyond the political motivation behind its lines, its objective is to prioritize freedom of expression, even if it causes harm to other fundamental rights of equal importance, such as privacy, honor, image, and accurate information.
In order to do so, the provision of its rules seeks to excessively interfere with the self-regulation power of social media managers and related applications, limiting the moderation of content and accounts only when the “justifiable causes” defined in a restrictive manner in the project are established, which disproportionately violates the fundamental right to free initiative.
In this line of thought, we conclude that Bill No. 3227 of 2021, just like the content of Provisional Measure No. 1068 of 2021, is materially unconstitutional and, therefore, its processing or approval should be denied by the National Congress.