Some information about these messages, such as their authors, has also been made available. A study by Privacy International, in partnership with InternetLab and ELSAM (2021), points out that the mechanisms in place are still not very efficient, and that there is a lack of clarity about the countryspecific criteria for regulating advertising; there are also disparities among nations with respect to the timing of implementations. One of Facebook’s measures that generated controversy was the establishment of exceptions and greater tolerance for political leaders. This occurred in several countries, such as the USA and also Brazil. After several posts and live broadcasts in which Brazilian President Jair Bolsonaro contradicted the guidelines established by health authorities, Facebook and Instagram took down the live broadcast in which Bolsnonaro compared Covid-19 vaccines to contracting HIV/AIDS (Souza 2021). Facebook has established a purportedly independent oversight board, comprised of researchers from different fields and equipped with the prerogative to alter the moderation measures taken, for the purposes of analyzing cases that are deemed complex. One of the first decisions announced in 2021 pertains to a post by Brazil’s Federal Council of Medicine, in which it spoke against lockdown as a measure to combat the pandemic. The oversight board upheld Facebook’s decision not to take any measures against this post. During the review process, the Brazilian organization InternetLab recommended measures such as the application of a disinformation label (Péricles and Santos 2021). WhatsApp, a messaging app also controlled by Meta, has invested in measures to limit the potential for messages to go viral, such as restricting the forwarding of messages that are being reposted frequently. The company has also established partnerships to deal with content disseminated by electoral or health authorities. Studies such as those by Chagas, Modesto, and Magalhães (2019) and by Santos et al. (2019) illustrate how the platform was heavily used to disseminate disinformation during the 2018 Brazilian elections. Following a report in the newspaper Folha de Sáo Paulo (Mello 2018) on the dissemination of bulk messages containing false content during the elections, WhatsApp took down bot accounts engaged I N FO R M ATI O N D I SO R D ER IN THE GLOBA L SOUTH in bulk messaging (Valente 2018). YouTube has adopted a policy to reduce the spread of harmful disinformation and what it calls “borderline content.” Channels that falsify their identity and manipulate material (i.e. by editing it to mislead users) are also banned. The platform has begun removing content that violates its policies on elections and, more recently, on the Covid-19 pandemic. It has also announced that it is offering contextualized information from “authoritative sources.” However, after banning the monetization of videos associated with disinformation during the pandemic, the company has gone back to allowing this practice, collecting and paying content creators who have profited off of disinformation (Yahoo!Noticias 2021). In its response to a survey that was conducted as part of this research, Twitter affirmed that its focus is not to determine whether information is false or not, but rather to curb the potential harm of misleading or deceptive content. The platform’s actions include taking down fake accounts, removing posts that contain certain types of disinformation (e.g. health-related), and offering context along with information. The platform also has a policy to restrict manipulated media, such as deepfakes. “With respect to preventive solutions, we have sought to disseminate information about media education in partnership with nongovernmental organizations in this field and more broadly in the field of human rights,” the company said in its answers to the survey. A study by the Brazilian organization Coletivo Intervozes (Barbosa, Martins, and Valente 2021) analyzed the measures undertaken by platforms to counter disinformation and concluded that these measures are poorly structured, insufficient, and generally adopted because of great societal pressure and pressure from public authorities. The study also found that these companies don’t generally have their own structures in place to deal with this issue, their rules are unclear and hard to find, and they don’t adopt substantial solutions that get to the root of the problem, which is in fact associated with their business models. The release of Facebook’s internal documents in October 2020 and investigations by media outlets have shown that, in terms of combating disinformation, the company’s focus, which is already weak, has been on English-language content. An investigation by The Washington Post (Valencia 2021) showed how not just Facebook but also other platforms like YouTube have dealt feebly with the issue of disinformation in Spanish, which Latin America and the Caribbean – 9 1