The Tale Of Two Cases (And Platforms): The Brazilian Paradox

Background of the law

In Brazil, the liability of internet intermediaries such as Google, more commonly referred to as 'platforms', is provided under the law of Marco Civil da Internet (hereinafter “Marco Civil”). This safe harbour law was introduced by President Dilma Rousseff's government in 2014. It makes a basic distinction of intermediaries as Internet Connection Providers (ICP) and Internet Application Providers (IAP). An ICP would refer to any broadband service provider. An IAP, in this context, is any platform dependent upon the Internet to provide a service like a search engine or a social media website.

Within the Marco Civil framework, Article 19 simply provides that an IAP can only be held liable for any illegal content on its platform, if it does not fulfil an explicit court order to remove the said content. One exception is the notice-and-takedown regime envisaged under Article 21, allowing the IAP, to make unavailable any content, without a court order, simply at the user's request. This reliance on a user notice can only be justified if the matter concerns a compromise to their privacy (to mean 'nudity' or 'sexual acts of a private nature' only). Also, Article 19 does not apply to copyright infringement, as contained in the Brazilian copyright law.

Supreme Court's (Perceived) Dilemma

However, the two cases brought before the Supreme Court may change the above status quo. These matters concern fake profiles of users created by mischief-makers. The affected users attempted to approach the concerned intermediary to take down the profiles based on Article 21. Yet, the platform did not remove the said content as they do not fall under the categories specified under the said provision. Further, there was no reason for this entity to act under Article 19 as there was no court order. Therefore, the users/petitioners have challenged the constitutional validity of Article 19. The lower court asked the concerned platform to remove the said content, after declaring the provision unconstitutional. Further, the court reasoned that the imbalance created by Article 19 between freedom of speech and expression on the one hand and right to privacy of the users and removing 'harmful' content on the other hand makes it unconstitutional.

For a deeper understanding of the implications of this constitutional strikedown, let us appreciate the methodologies of removing illegal subject matter on the Internet.

Usually, intermediaries receive a 'notice' based on which they 'takedown' content. It is this 'notice' that has been subject to scrutiny. It could be at a user's request (like we see in the USA) or on the basis of a court order.

Courts adjudicate matters and are bound to provide appropriate reasons for their judgment. Due to this inherent nature of court orders, freedom of speech and expression receives primacy and priority. However, in the 'speedy' world of Internet, an intermediary or a platform acting upon a user's request, or screening content on its own, is time-saving and efficient, even at the cost of censoring genuine content. To give a copyright-based illustration, a study by CREATe reveals that, of the 1,839 parody videos studied, an alarming 32.1% of the content were taken down by the algorithm itself in 2016. Therefore, a total reliance on technology offered by the intermediary for removal of “illegal content”, while suitable for the Internet, could have human right and constitutional consequences.

Therefore, a court order, as propounded by Article 19, would ideally ensure no chilling effect in the takedown of content. An unwavering basis of court orders in the law may also arrest the temptation of an intermediary to monitor its own platform.

So, where is the paradox in this tale? Maintaining the validity of the concerns for privacy and reputation damage, the idea of striking a provision (which may be the key to protecting freedom of speech) on “constitutionality”, is intriguing. Rationalizing the observations so far, this paradox is born from the importance awarded to both right to privacy and freedom of speech and expression constitutionally.

This 'effective' censorship brings to mind the grounded observations of Shreya Singhal v Union of India. Striking down a broad-worded Section 66A of the Information Technology (IT) Act, 2000, the Supreme Court of India, reinterpreting Section 79, restricted the intermediaries from removing content on any other request except a court order. This was hailed by many as an exercise of constitutionality, preserving the essence of freedom of speech and expression until the IT Rules of 2021 were notified. Contrasting the Shreya Singhal observations, these Rules expect an intermediary to use 'technology-automated measures' to remove illegal matter.

Interestingly, some commentators of Marco Civil have disagreed with the 'freedom-friendly' interpretation of Article 19. They have opined that the provision is silent on prohibiting an IAP from monitoring content. Moreover, there isn't a customary “no general obligation to monitor” restriction/provision for IAPs, like in the EU (E-Commerce Directive and Copyright DSM Directive). Yet, Article 19, with its flaws, seems to have great potential.

Key Takeaways

From an understanding of the letter, Marco Civil's Article 19 could really benefit from adding a restriction against generally monitoring content and optionally, from expanding the meaning of privacy under Article 21, to include certain methods of misappropriation of personal information. Care should be taken to not restrict freedom of speech and expression.

For the Internet to flourish, it is necessary that the grounds and the methodology for removing content by the platforms is sound with the spirit of the law. There may be some flaws with striking down a provision and permitting an intermediary to entertain user requests and independently monitor content in anticipation of such requests. Defining 'privacy' and its grounds is crucial. In my opinion, it may still be more definable than 'obscenity' and 'integrity' (of the sovereign), as mentioned in the Indian IT Rules of 2021.

Technical prowess, as empowering as it may be, cannot be completely reliable. It has been seen that, intermediaries, for business and reputational interests, naturally levitate towards automating processes such as removal of matter through algorithms. Further, they have also shown a tendency to remove more than what is necessary (the toddler dancing to 'Let's Go Crazy' immortalized a YouTube copyright strike in our minds). In other words, the more encouragement the law gives to a platform to use technical measures, the likelihood of censorship and a compromised freedom of speech is higher.

The Brazilian matters are currently sub-judice. However, one could really contemplate the 'what-ifs' by brushing some clear strokes on Article 19 with no ambiguity on proactive monitoring, retaining its inherent constitutional nature. What can India learn from the Brazil example? Intermediaries and their prized algorithms are no courts.

The author is an Assistant Professor, Jindal Global Law School. Views are personal.