Why PL 2630/2020 is Dangerous
Este texto está disponível em Português
Democracy is under attack in Brazil. The Chamber of Deputies is soon expected to vote on PL 2630/2020 that was recently changed to include over 20 completely new articles that have never been widely debated. If it is passed, companies like Telegram may have to stop providing services in Brazil.
Here’s how this bill will kill the modern internet if passed with the current wording.
It Gives Censorship Powers to the Government
This bill allows the government to limit what can be said online by forcing apps to proactively remove facts or opinions it deems “unacceptable” [1] and suspend any internet service – without a court order. [2]
For example, the Minister of Justice recently called for sanctions against Telegram, claiming it “didn’t respond to a request” – before that request was even made. If PL 2630/2020 had been active, the government could have immediately blocked the app as a “preventive measure” until Telegram proved that it did not violate any law.
It Surrenders Judicial Powers to Apps
This bill makes digital platforms responsible for deciding which content is “illegal” instead of the courts – and gives them overbroad definitions of illegal content. [3]
To avoid fines, platforms will choose to remove any opinions related to controversial topics, especially those that do not correspond to the view of any government currently in power, which directly endangers democracy.
It Creates a Permanent Surveillance System
The bill requires platforms to monitor communications and inform law enforcement authorities in case of suspicion that a crime has occurred or may occur in the future. [4]
This creates a permanent surveillance system, similar to that of countries with anti-democratic regimes.
It’s Unnecessary
Brazil already has laws for tackling the criminal activities this bill aims to cover (including attacks on democracy).
The new bill aims to bypass that legal framework, allowing a single administrative entity to regulate speech without independent and preemptive judicial oversight. [5]
And More!
This only touches the surface of why this new bill is dangerous. That's why Google, Meta and more have all joined to show the congress of Brazil why the bill needs to be rewritten – but this won't be possible without your help.
What You Can Do to Change This
You can talk to your deputy here or on social media today. Brazilians deserve a free Internet and a free future.
References
1
PL 2630 establishes itself as the Brazilian law on liberty, responsibility, and transparency on the internet, but the current wording added on April 27, 2023 makes it potentially dangerous. While Article 3 outlines some principles to uphold freedom of expression and prohibit censorship, subsequent articles heavily undermine these protections. Namely, beginning with Article 4, the law imposes obligations that facilitate censorship by promoting far-reaching moderation practices on ambiguous, broad and sometimes undefined subject matters.
Article 33 extends parliamentary immunity of government officials to online platforms. In so doing, this provision establishes a dangerous imbalance in which the government receives protection while the speech of regular citizens or election candidates, especially on contentious topics, may be subject to arbitrary moderation.
Article 12 allows for a 30-day "security protocol" imposed by the government. The protocol can be extended based on arbitrary and ambiguous factors like "negligent" or "insufficient" moderation measures enacted by service providers. Under this protocol, providers are held liable for damages from third-party content deemed “unacceptable”, potentially encompassing a wide range of opinions or topics, as defined by the government. The security protocol effectively compels service providers to enact and align speech moderation with the government's preferred narrative, possibly requiring updates to their terms of service to ban specific subjects or viewpoints altogether.
2
Under Article 47, a service may be suspended in an administrative procedure, without prior consultation with independent courts. The suspension can also be done as a “preventive measure”, before the procedure even begins. While it is, in theory, possible to file a lawsuit seeking to overturn such a decision, the service may remain suspended throughout the entire time that it takes the government to conduct the administrative procedure (and then for the court to reach a final decision). Additionally, if PL 2630/2020 is passed, the court would have to base its rulings on the same vague and overbroad definitions that allowed for the initial suspension (see note [3]). Even if the court ultimately rules that the suspension had been unlawful, by then the service could have stayed blocked for months or even years.
3
Under Article 11 §1, an app can become liable for content as soon as it gets reported by any user, unless the app proves that it handled the notification diligently. Under the “security protocol” (Article 13), this liability becomes immediate for any content related to the protocol, based solely on user reports.
As a result of the due care obligation and the need for mitigating “systemic risks”, platforms are obligated to proactively monitor the entirety of their services for “illegal content” that is defined as relating to a (potentially non-exhaustive) list of laws – which includes “crimes against democracy” among other things. However, the listed laws are intended to regulate conduct, not speech – as such, they are not suited for handling opinions expressed online in the way envisioned by the bill.
For example, under the current wording of PL 2630/2020, the government is free to declare almost any content touching on the three branches of government (e.g., criticism of the Supreme Court or President) a crime against democracy that must be banned under risk of app suspension.
4
The requirements to analyze and mitigate systemic risks (Article 7) and act with due care (Article 11) can only be met by monitoring and policing a substantive amount of (not to say all) users’ content and creating an aggressive surveillance system. Under general transparency and auditing obligations, information about users’ activity can be requested and accessed by the government.
Also, specific reporting obligations on any suspected life threats that are set by Article 45 might lead to over-referring of any aggressive speech on the platform to law enforcement authorities. While the text mentions that platforms must only look out for "information that raises suspicions that a crime involving a threat to life has occurred or may occur" – requiring platforms to proactively seek out information of any kind necessitates that the platform must therefore monitor information of all kinds.
5
Article 11 contains a list of laws that currently cover the intended scope of PL 2630/2020. These already existing laws govern crimes against the democratic rule of law, terrorism, racism and violent or discriminatory behavior, among others. The presence of these comprehensive regulations indicates that PL 2630/2020 would only serve to undermine the established legal framework, potentially granting the government excessive power and negatively impacting the delicate balance between security and personal rights.
The references are based on PL 2630/2020 as published on April 27, 2023, available here.