10 ways to combat disinformation

How do we combat disinformation without weakening freedom of expression?

Disinformation produced with the intent to cause harm is not a new phenomenon, but has taken on a new form, due to widespread access to connections and mobile devices, the popularization of various means of media production, concentration in a handful of global platforms that support content generated by third parties – whose business is based on data collection, micro-segmentation marketing and maintaining their audience – and content migration to instant messaging systems, such as WhatsApp and Telegram.

Given such a complex phenomenon, and one with such a new profile, it is true that few certainties remain, nevertheless, it is important to assert them.

It is necessary to combat disinformation with a view to strengthening the public debate arena, in other words, without jeopardizing the right to access information, freedom of expression, privacy, and the due process of law, while it is also essential to ensure adequate punishment for violations and crimes, and to promote the right to reparation.

The Joint Parliamentary Commission of Inquiry (Comissão  Parlamentar  Mista  de Inquérito: CPMI) into Fake News is working within the National Congress with the prerogative to investigate the creation of fake profiles and cyber-attacks aimed at influencing last year’s elections and prejudicing public debate.  It is therefore important to understand the risks and opportunities of including legislative recommendations in its final report.

 

What to avoid?

The error lies in the choice of object for regulation (the user) and the strictly penal treatment, which may place a greater burden on the legislature and the prison system, and lead to censure and self-censure – since there is no consistent definition of disinformation which may be applied without taking an infinite number of factors into consideration.

According to Article 19 of the Civil Rights Framework for the Internet, the application service provider will only be held civilly liable for damages arising from information generated by third parties following a judicial determination to take a specific action which is not attended to within the specified period. This rule was created to guarantee legal security for the platforms, meaning that they are not obliged to proceed with an illegitimate removal request, and therefore protects the exercise of freedom of expression.  This does not mean that the platforms cannot enforce moderations in line with their policies (as long as these do not conflict with Brazilian legislation). Such moderation may occur, but if the decision about something so complex is imposed by law, rather than undertaken by the State, it falls to the responsibility of private companies, expanding the risks of restricting freedom of expression and reducing the plurality of content available online.  It is worth remembering that, under current law, cases related to honour, reputation or personality rights may be brought before special courts.

It is worth considering ways to speed up processes on the part of the judiciary – which, indeed, is very slow – to analyse content intended to cause harm, such as strengthening civil lawsuits and therefore meeting international standards to protect freedom of expression.  These suggest that intervention measures related to the circulation of information should come under the responsibility of judges, trained for this purpose, prepared for an analysis that balances freedom of expression with other rights, and whose decisions are subject to public scrutiny and analysis by higher bodies.

So, what should be done?

Intervozes understands that it will only be possible to construct a healthy flow of information for citizenship and democracy through a series of measures.  There is no magic solution to the problem of disinformation; we need to confront it in its complexity. Some suggestions are listed here.

 

1)  Punish the artificial and industrial use of instant messaging platforms

Behind disinformation campaign strategies on WhatsApp are companies using automated and bulk messaging methods.  Sending bulk messages based on automated systems is a practice prohibited by the Superior Electoral Court (Tribunal  Superior  Eleitoral: TSE), but progress is required in the identification and punishment of these companies.

During the electoral period, WhatsApp, a company in the Facebook conglomerate, announced it was banning hundreds of thousands of accounts from its application.  These profiles are suspected of sharing fake news and were tracked with the help of a filter that automatically identifies spam. According to the company, the list includes numbers of agencies that sell bulk messaging, but so far none of these companies has been investigated.  Why?

There is no legislative proposal that can lead to a healthy information flow if crimes are not investigated and those responsible duly punished.

 

02) Right of rapid and proportional reply

To combat disinformation, it is also necessary to seek speedy instruments to repair damage, given the scope and speed of the distribution of messages that trigger the guarantee of the right of reply, and enable any reply to reach the widest, potentially affected audience.

When granting the right of reply, the judiciary must take into consideration the scope and speed of publication of content that motivated the request for redress and propose measures that have widespread repercussions.  The dissemination of denial or clarification messages should reach users affected by the misleading content at the same speed or even, in specific cases (such as in the final stretch of an election), more quickly.

 

03) Protect personal data and monitor its use

Much of the operation to distribute in bulk form disinformation intentionally generated to cause harm is based on the use of personal data.  The General Data Protection Law (Lei Geral de Proteção de Dados: LGPD) (Law 13.709/2018) was approved precisely to guarantee the rights of data subjects, create legal bases for this treatment and monitor the use of this important asset in the information age.  If the LGPD were respected, the effectiveness of disinformation practices could be reduced and, thus, discouraged.  One of the measures recommended by the NGO Safernet is the prohibition of the use of the target audience’s profile to direct advertising through promotions, including any information related to sensitive aspects such as racial or ethnic origin, religious beliefs, affiliations to unions or organizations of a religious nature, data regarding health or sex life.

It is necessary, however, for the National Data Protection Authority  (Autoridade Nacional de Proteção de Dados: ANPD) to have the political autonomy and financial independence to undertake the work of education, monitoring and sanction of the law.  It is also worth updating the electoral legislation to adapt it to the LGPD, giving citizens control of the way that parties and electoral campaigns currently use their data.

 

04) Prohibit advertising by candidates and online campaigns

Twitter recently announced that it would no longer accept political advertising on its platform.  Advertising paid for by candidates and political advertising as a whole will be restricted.  This is because Twitter understands that political adverts on the internet present entirely new challenges to civic discourse: optimization based on messenger machine learning and multiple segmentation, unverified misleading information and deepfakes – undetectable montage.  All this with increasing speed, sophistication and overwhelming scale.  If Twitter can do this, why don’t the other platforms follow its lead?

In the same way that during the age of broadcasting it was determined that campaign insertions would be governed by rules in order to provide visibility proportionate to the size of alliances, rather than to the candidates’ economic power, the rule for the internet cannot be that he/she who pays the most, wins.  The aim of internet advertising paid for by candidates will also have a positive impact on the distribution of paid disinformation online.

In 2018, the organization SaferNet presented 14 proposals related to this issue to the Superior Electoral Court (Tribunal  Superior  Eleitoral: TSE).  These included the prohibition of so-called hidden posts or dark posts, paid posts directed at a specific audience which the rest of the population cannot see; and prohibition of payment for adverts and promotions containing political content in foreign currency, according to a document delivered to, but not assimilated by, the court.

We would add the need, during the electoral period, for platforms to be open with the TSE about publications with a greater audience reach, including those that are not promoted, which mention candidates, coalitions and parties, so that the body can proactively analyse the need to take measures, such as offering the right of reply or suggesting restrictions to the reach of certain content.

 

05) Integrated access to internet applications

Today most Brazilians access the internet through broadband on mobile phones and do not have continuous data packages. In Brazil, therefore, internet use is limited to a few applications, through so-called zero-rating (or zero tariff) agreements, which the telephone providers enter into with certain platforms.  In this model, navigation ends up restricted to these applications, so that any information checking  or the reading of an article in its entirety on the website of origin, for example, are compromised, strengthening the phenomenon of disinformation through images and headlines, which often use arresting strategies in search of clicks, without presenting the facts in their complexity.

The containment of disinformation therefore necessitates expanding access to quality connections at affordable prices and to the whole internet. To this end, one important measure is the prohibition of navigation-blocking once mobile data has been used up, guaranteeing the Mobile Network Service and Multimedia Telecommunications Service providers the freedom to reduce navigation speed, provided this does not invalidate the right to access.

 

06) Transparency and autonomy in relation to algorithms

It is essential to consider the effect of new content distribution models on the construction of the public sphere.  In this sense, the impact of algorithms to filter and prioritize content when searching for news and information on platforms is significant.  Bias in choosing information has always existed, but the potential that exists today for a citizen to be unconsciously confined to a specific spectrum of political information is completely different and creates a bubble effect, one of the negative, unforeseen and unintentional results of this business model.

To repair damage, it is important to allow the user to choose the content they visualize (in news feeds, search results, news access services and the like) and be transparent about the criteria algorithms use for ordering and/or targeting, if possible explaining effects to the user.  Large platforms must provide custom filtering mechanisms in a transparent, revocable/editable manner, which is under the user’s control, so that it is the user who finally decides the content they wish to prioritize and they do so (for example, they may prefer to see content posted by chronological order rather than based on the definition of their profile), while activities to prioritize online content should be accessible to the user (feeds, search results and others). Commercial agreements that interfere in the organization of content must be visibly displayed alongside content.

It is also essential that positive measures are taken to promote diversity and plurality on platforms, enabling the use of algorithms to expand communication flow, as opposed to the current situation, in which, due to algorithm prioritization, certain content does not have the desired internet reach.

 

07) Breaking digital monopolies and oligopolies

As with traditional media, the concentration of digital media prevents diversity and plurality in public debate, weakening democratic societies.  In this sense, given the specificity of communication, because of its symbolic (cultural/social/political) nature, activities to promote competition and antitrust policies need to be allied to theories and practices to protect and promote freedom of expression, pluralism and diversity.

Given this understanding, regulators and the authorities responsible for the application of competition and for public communication policies should develop ex-ante analyses and actions to predict the impact of economic concentration on diversity and plurality.  Once markets are defined, it is very hard to correct asymmetries, which end up becoming structural.

Restrictions on horizontal, vertical and cross- ownership should be introduced for agents with significant market power.  In the case of large platforms, where the same agent has several businesses, these should be broken up.

 

08) Incentives for responsible, plural and diverse journalism

Diversity of voice and position is an essential factor for the health of democracy.  It is therefore necessary, as required by the Constitution, to establish rules to prevent media monopolies and oligopolies and develop mechanisms to encourage the production of responsible content aimed at promoting the public interest.  Further, regarding the application of public funds to media outlets, it is essential that objective and widely disseminated rules be established that consider not only audience, but also diversity and commitment to the public interest, in line with Article 221 of the Constitution.  Although these guidelines are intended for public communications, they should also guide the investment of public funds in the media sector in general.

 

09) Media education

Finally, it is necessary to create instruments to promote a critical analysis of traditional and new media by the population as a whole and on a large scale. Although schools are certainly privileged arenas for training, they are not the only ones.  It is worth noting that any media training needs to go beyond an analysis of content and address the economic structure of the media in Brazil and the world, as well as the country’s regulations and other jurisprudence. This is the only way to train citizens to be critical of the information flow, taking account of content and of the political and economic interests of media companies, something not necessarily visible in an individual analysis of text, video and other productions.

Specifically, based on multi-sectoral participation, it is necessary to develop a proposal for a wide-scale educational process that involves the media and schools; to incorporate media education into the school curriculum, based on the premise of the school as a locus for reflection about what actually happens in people’s lives; as well as to support, through public policy activities, media education developed by civil society organizations from diverse territories.

 

10) Transparency in server communication

Due to the absence of procedures that guarantee the storage of public server communications and their periodic availability for public oversight, it is worth discussing the need for the regulation of the official communication channels available to servers, as well as for processes to safeguard and store these records and subsequently make them available for public access by Brazilian society.

 

Source: Intervozes (www.intervozes.org.br)