Powerless against Digital Hate - International Reports
This portlet should not exist anymore
The minister is here in a private capacity. A prominent journalist in Mbabane told him that there are German visitors in the country. In the restaurant of a good resort in Manzini, over a rock shandy, a drink mixed from Angostura, soda, and lemonade, he explains that he knows this hotel. During the unrest of mid-2021, when barricades were burning in the small Kingdom of Eswatini, he spent two nights here. ‘Was forced to spend them’ would be more accurate, however, since his private house is not far away. On Facebook, he says, there was so much agitation against the king, the elites, the ministers, and state officials, that he had fears about his safety. Facebook, he says, shaking his head in disbelief, is a major force in Eswatini. Remembering the hate on the net during those days makes him shudder.
Social media are changing politics, even in the last absolute monarchy in Sub-Saharan Africa. In the small kingdom once known as Swaziland, where everyone seems to know each other, rumours spread like wildfire, especially on Facebook. Opponents and critics of the king deliberately spread rumours, be it about the king’s style of ruling, or his anti-China policy. Many of the posts are made from outside the country, especially in South Africa, where a number of critics of the royal house live for security reasons. An entire country can be destabilised by a social network; there is no room for debate, and outrage and agitation dominate. But what the Eswatini minister said about the threat posed by Facebook would be difficult to articulate in a public discussion because it would be perceived by some as an implicit attempt at censorship and restriction of freedom of speech.
Indeed, social networks of course have an emancipatory potential, especially in an authoritarian country such as the Kingdom of Eswatini. The king’s critics are able to effectively voice their criticisms for the first time. Years ago, the opposition might have smuggled flyers into a country; today, criticism flows in far more easily, through digital means. In recent years, rulers from Tanzania to Uganda to Zimbabwe have, therefore, greatly restricted the right to free speech by such parties as bloggers under the guise of combatting cybercrime. Rwandan President Paul Kagame, however, is doing something different: all networks are allowed in the Central African country, which regularly draws appreciative comments from Western visitors. Yet no Rwandan in the country will publicly quote Kagame’s critics, even if everyone has read them. The people know what is possible – and what must be avoided at all costs.
Corporations with Tremendous Reach
According to a new survey by research institute Afrobarometer, 92 per cent of people in African cities have a mobile phone. In rural areas, that figure reaches 76 per cent. Of the urbanites, 69 per cent have regular access to the internet, and 60 per cent get their news from social media. In rural regions, where the majority of people in Sub-Saharan Africa currently live, only a quarter of the population has regular access to the internet, since coverage is often inadequate and incomes insufficient for a smartphone or data package.
Still, 3.6 billion people regularly use a Facebook product, or Meta, as the company is now called. That is half of the people on earth. More than 80 per cent of Facebook and Twitter users live outside the US. In December 2020, there were 233 million Facebook members in Sub-Saharan Africa, i.e. about a quarter of all inhabitants. No other medium has such reach across languages and ethnicities. Neither the African Union, nor any UN organisation, to say nothing of any government or company, has such direct access to Africa’s people – and to their data.
The negative consequences of social media, especially of Facebook, are an existential threat to some. For instance, Arrey Elvis Ntui of the International Crisis Group lamented in February 2020 that ethnic conflicts in Cameroon were being exacerbated on Facebook with its four million users. In the Central African country, Facebook is “used to spread ethnic stereotypes, exchange ethnic-based insults, propagate misinformation and incite violence”. The network’s algorithms heighten tensions. Some observers think that this is not an unfortunate error, but a conscious strategy.
Unlike Europe, Africa rarely questions the omnipotence of social media. In an exchange of open letters between Springer CEO Mathias Döpfner and EU Commission President Ursula von der Leyen in early 2021, the necessity of limiting the power of tech companies was debated. Döpfner warned above all against surveillance of citizens, although the preservation of economic power of media companies, like the one he heads, was certainly also a (legitimate) concern. Von der Leyen spoke of the necessity of “limiting” the power of tech companies. But such discussions remain remote from Africa and its realities.
The option of ignoring Facebook, or even leaving it behind, is a Western privilege, as digital expert Rosemary Ajayi explains. For many people in Africa, the services of Facebook, Instagram, and especially WhatsApp have no alternative – not least because they are so cheap. In Africa, too, Facebook portrays itself as the great contact machine that brings people together who have lost touch with each other or who find each other on Facebook in the first place. The network presents itself as very modern and hip, and the African Facebook headquarters in the Bryanston suburb of Johannesburg sells itself in press and PR reports as people-oriented, a place where people get together in modern conference rooms with melodious names such as Tshongololo and Timbuktu. Facebook’s failings and the great security threat it poses to already fragile African states are not welcome topics of discussion. Kojo Boake, who heads Facebook’s Africa Policy team in London, explicitly declined an interview for this article with the Konrad-Adenauer-Stiftung’s Sub-Saharan Africa Media Programme.
Meta’s secrecy, smacking of sectarianism, may in individual cases make its power appear greater than it actually is. But Meta CEO Mark Zuckerberg has set global digital dominance as a goal. This involves making Facebook and Meta products accessible even to those who otherwise would have no access for technical or economic reasons. That is why Facebook established the Free Basics Initiative. Users download an app that gives them access to a few selected sites. Wireless providers charge no fees for the app. Users who want to access sites not included in the app are asked to buy data. What remains opaque is who has selected what internet sites for the app and why. In June 2020, the Free Basics Initiative was running in 32 countries in Africa. Kenyan author Nanjala Nyabola says that Facebook’s power is growing in Africa, but it is happening under the radar. Yet, she notes, “Zuckerberg doesn’t get Africa after all”. Many telecommunications companies cooperating with Facebook in the Free Basics Initiative are government-controlled or partially government-owned. This means that Facebook chooses which news or products a user in a slum in Nairobi or Abidjan receives. The system influences purchasing decisions, but also political attitudes towards events in the users’ area. For instance, if one searches for “Kenya Airways” online, one reads critical reports about the troubled airline, its poor service, and constant delays. Yet the same search through Free Basics returns only the airline’s website and flight offers. In India, Meta’s Free Basics Initiative failed and had to cease operations. What The Economist wrote about Facebook also applies to the initiative: “a few unelected and unaccountable executives are in control”.
Authors Sheera Frenkel and Cecilia Kang warn that Facebook is “the world’s biggest testing lab, with a quarter of the planet’s population as its test subjects”. It could become a “digital Frankenstein”. One of the biggest problems is that Meta’s executives are not even aware of what is happening in other languages on their networks.
Awareness of the Problem Remains a Challenge
In Africa, too, awareness of social media’s dangers is not as great as satisfaction with the opportunities it offers. At first, there was very great joy, including among intellectuals, that the bonds of censorship were loosening. Botswanan media manager Joel Konopo wrote that “It’s difficult for African governments to ignore 500 million liberated voices.” Nic Cheeseman of the University of Birmingham recalls that it was taken for granted that new technological capabilities would strengthen democracy. Today, he says, there are far more pessimists. “WhatsApp is a disruptive technology that challenges existing hierarchies in ways that are simultaneously emancipatory and destructive.” Thus “[t]he emancipatory effect of WhatsApp will be most significant in authoritarian countries marked by repression and censorship”. Its “destructive potential […] is likely to be greatest” in countries that are fundamentally democratically oriented, but in which “public trust in political institutions [is] low”. A study conducted by the University of Johannesburg and Africa Check on the effects of WhatsApp shows that encryption leads to greater secrecy, not least in the form of forwarding false information, which is then not corrected despite better knowledge, resulting in a vicious cycle.
Most African rulers have fewer problems with Facebook and other Meta products than with their neighbours, says Justine Limpitlaw, a media law expert at the University of the Witwatersrand, South Africa. The media law situation is worst in Eritrea, she says, since citizens there do not even have access to the internet. The situation in Ethiopia has also deteriorated dramatically. Kenya, on the other hand, impresses her. South Africa is doing relatively well, with the best data protection laws on the continent, although the security services simply do not apply these useful rules.
Due to persistent fears of an African version of the 2010 “Arab Spring”, most Sub-Saharan African governments are primarily interested in surveillance. The need to protect citizens from cyberbullying, internet fraud, and fake news is largely ignored. Instead, social media activities are impeded to a very great degree by such measures as a social media tax in Uganda and mandatory blogger licensing in Tanzania.
Politicians in Africa do not necessarily understand how social media work. But they know that they need it. Nigerian President Buhari was affronted by Twitter when one of his tweets was deleted because of a derogatory comment about the Igbo ethnic group. Buhari’s comment was reminiscent of the Biafran War of the 1960s. The reaction of the president of the most populous African country was not long in coming: access to Twitter was blocked in Nigeria. Besides appearing infantile, the move caused a surge in virtual private network (VPN) downloads in Nigeria. Such secure channels allow access to blocked sites, such as Twitter.
The American social network Twitter, which has been successful at giving itself a much friendlier image than Facebook, but which nevertheless allows its algorithms to make some fairly questionable decisions, installed a team in Ghana with the justification that the country protects freedom of speech and guarantees freedom on the internet. Meanwhile in 2021, Twitter blocked an issue of The Continent, a pan-African WhatsApp newspaper, because it contained Bill Gates’ name and discussions about his medical programmes in Africa. The newspaper had included an interview with Gates; however, the algorithm had simply detected criticism of Gates and blocked it – with predictable consequences for sales and advertising of the issue in question. Since Elon Musk announced the takeover of Twitter, there has been a great amount of insecurity, also among Twitter employees, who have since declined to talk to critics, saying they first have to wait and see how the network’s policies develop.
Many academics and journalists on the continent who are trying very hard to limit the negative effects of social media also believe there is ultimately no way of combatting them. In 2019, the Konrad-Adenauer-Stiftung’s Media Programme was invited to Addis Ababa to discuss a new Ethiopian media law with German experts and Ethiopian jurists, editors-in-chief, and university lecturers. Abiy Ahmed had recently become prime minister of Africa’s second-largest country. He had released jailed journalists and bloggers, and promised that everything would get better. The discussion in Addis Ababa was about the Ethiopian and German situations, which were certainly difficult to compare because there was very little printed media in Ethiopia, on top of which the internet had been repeatedly controlled, and even shut down. Moreover, many heads of Ethiopian media companies saw their role primarily as representatives of their own business interests, rather than warriors for truth.
Digitally Fanning the Flames
A constant throughout the discussions was the significant concern regarding the uncontrollable power of the tech giants. In the Ethiopian Highlands, Facebook and YouTube in particular are not only used for posting shopping tips and wedding wishes, but also for political and ethnic agitation. Gangs from the two largest groups, the Amhara and the Oromo, use Facebook to plan brutal actions against one another.
Prime Minister Ahmed used to work in military intelligence, where he was responsible for controlling the internet. Several times he had shut it down completely in order to (officially) calm the situation in the country. Yet none of this changed the feeling of paralysis on the part of the political and media class, of being completely at the mercy of uncontrollable forces.
A Google manager, at the 2019 International Press Institute World Congress in Geneva, when asked how many moderators Google had who could speak Amharic, appeared perplexed and replied that there were none as far as he knew. In the meantime, that may have changed for some social media giants. But the fact remains that content moderation is often used as a fig leaf and remains insufficient for the gigantic mass of racist or inciting content requiring moderation.
Facebook’s moderation of posts from embattled parts of the world, such as Ethiopia – so far removed from the artificial reality of Silicon Valley – continues to be inadequate. This is compounded by a lack of sufficiently qualified employees with historical and political knowledge, but especially of psychological understanding and intercultural abilities. It sometimes appears, as authors Frenkel and Kang say in their book “An Ugly Truth”, that the world is getting to be too much for Facebook and its more than 30,000 employees. Moreover, an apolitical boss like Mark Zuckerberg models to employees that politics and whatever happens out there in the world is fundamentally unsavoury, and in any case far removed from the reality in the US.
Zuckerberg and manager Sheryl Sandberg have publicly expressed regret that Facebook failed to confront, moderate, or delete posts calling for the murder of Muslim Rohingya in Myanmar. The Economist reports that, since 2010, Facebook has failed to moderate its content actively and effectively. Indeed, Facebook always regrets when a new security breach arises – be it when Russian internet trolls influence the American elections, or when Holocaust deniers propagate their crude theories. Yet Facebook’s algorithms appear to favour such uproar, up to and including calls to murder Muslims; it all seems to be part of the business model. After all, the greater the number of people viewing the image of a dead Burmese, the greater the potential business success.
Myanmar is a long way from Palo Alto. At least as far as the Ethiopian highlands. The civil war led by Prime Minister Abiy Ahmed against his former allies from Tigray, is being fought not only in Lalibela and other holy sites of Ethiopian Orthodoxy, but also on the internet, especially on Facebook. After viewing documents by whistleblower Frances Haugen in 2021, The Continent reported that Facebook is intentionally failing to do what is necessary to take such measures as deleting calls for murder. “[I]t appears that the prevalence of hate speech and disinformation on Facebook is not a bug. It’s a feature,” say the Continent authors. An official information request by the Konrad-Adenauer-Stiftung’s Media Programme to Kojo Boake, Facebook’s Public Policy Director in London, asking, among other things, how many moderators speak Amharic, received the answer that Facebook is not in a position to respond to the request at this time.
Rosemary Ajayi of the Digital Africa Research Lab says that “Facebook’s poor handling of the Ethiopia crisis” should be a warning to the rest of Africa “because it is indicative of the quality of response we are likely to see in other African countries”. In October 2020, Facebook proudly announced that it had hired more moderators who spoke Swahili, Amharic, Zulu, Somali, Oromo, and Hausa. But the reality of content moderation does not seem to reflect this. The case of Ethiopia is just one example of the company’s paternalistic treatment of other world regions. Research by Time magazine in early 2022 showed that moderation of controversial content is outsourced to external companies. Time reported 200 young men and women working as moderators for eleven different African languages in Nairobi. They view brutal videos and messages for Sama, a Kenyan company based in California, which in turn is commissioned by Facebook to moderate content on the African continent. The working conditions in Nairobi described by Time suggest that, in addition to trauma that occurs from constant viewing of violence, the sheer volume of posts to be assessed makes the work impossible. The company in Nairobi apparently assesses the lion’s share of critical posts from Africa. Just imagine: there are more than 250 million Facebook users in Sub-Saharan Africa, where Facebook remains the most important social media company, although the American company is likely to lose this position to TikTok over the next few years. And their posts (many users post multiple times per day) are being checked by just 200 moderators?
After each scandal that influences an election or costs lives, Mark Zuckerberg and others promise that more money will be invested in security. For instance, in 2021, Facebook is thought to have spent five billion US dollars for security measures, and hired 15,000 more moderators worldwide. Facebook does not provide exact numbers, and it must be assumed that the mass of the moderators work with English-language posts, primarily in the US and Western Europe, where pressure from legislators and governments is particularly great to limit the company’s power or, as proposed in the US, even to break the company up. That is also why investigators are looking especially hard, and why Meta is working especially hard at its explanations, moderation, and lobbying. In Africa, the company seems not to think that such efforts are necessary.
Many African governments fear critical bloggers and influencers more than societal disintegration due to social media. There is scarcely any political awareness of the dangers. In a country like South Africa with a broad middle class and relatively high standards of education, social media posts feature more racism and sexism than those of any other country on the continent. The Johannesburg Holocaust and Genocide Centre and the University of Cape Town are conducting a study on narratives and recurring stereotypes in posts. Dr. Gavaza Maluleke and Professor Adam Mendelssohn are heading a team of coders representing all of South Africa’s skin tones and social classes. A white person from the South African upper class does not necessarily understand the connotations of expressions used in poor districts, and vice versa. Besides the fact that many of the very dedicated coders report that the work concerns them and causes them to reflect on their own prejudices, many multipliers in South Africa are surprised at the number of posts that are justiciable. Yet the widely praised South African justice system is overwhelmed by the sheer volume. Still, sometimes a judge will sentence someone who denies the Holocaust or posts racist comments to a few days of reflection at the Holocaust and Genocide Centre in Johannesburg, where director Tali Nates, daughter of one of the Jews saved by Oskar Schindler, explains not only the Shoah, but other crimes against humanity such as the Armenian genocide, apartheid in South Africa, the 1994 genocide in Rwanda, and the Srebrenica massacre.
Who Can Stand Up to Facebook?
It is of course not only in Sub-Saharan Africa that tech giants feel invincible. The Australian government’s dispute with Google and Facebook about dissemination of, and potential payment for, Australian and other media content is an example. While the tech giants may have got a bloody nose in the process, it will not be their last attempt to disseminate for free content that others pay for. But there is not a single government on the African continent that would stand up for the interests of their citizens and media companies as the Australian government did. That makes it much easier for the large social networks, and they can be more cavalier and ruthless than, for example, in Australia.
Mark Kaigwa, head of Nendo, a digital consultancy in Nairobi, views worries about citizen data security as a Western privilege. “These worries about security on the internet are only for those who can afford larger quantities of pre-paid mobile data,” he notes. And that is just where companies like Facebook and TikTok come in: a majority poor population desirous of participating in the world, coupled with African rulers whose interests do not lie in protecting their citizens from cyberbullying or fake news – such a situation presents these companies no obstacle. If a mobile subscriber has no other currency than sharing their own data, then of course that is the form of payment they will use.
The timing of when Facebook will be seen in Africa, like it is in Europe, as the network of the older generation depends largely on the further development of YouTube and TikTok. For Kaigwa, they have the greatest potential on the continent. Writing in African Business, Will McBain outlines TikTok’s advantages: “African governments’ cosy relationship with China may protect [TikTok] from the occasional social media bans that blight the continent’s digital landscape.” The company was founded in 2016, and by April 2020, its app had been downloaded more than two billion times. TikTok’s management considers Africa its largest growth market, spurred by increasingly affordable internet-enabled smartphones made in China. Beijing has purchased a one per cent share of Byte Dance, TikTok’s parent company, giving it a seat on the board, which has only three members. In 2020, Statista identified a TikTok market share of just under 32 per cent, with technical centres in South Africa, Nigeria, and Kenya. As an explanation for TikTok’s success, McBain quotes Scott Thwaites, Head of Emerging Markets for TikTok Global Business Solutions: “[V]ideos are recommended based on what – rather than who – you like.”
Is it a coincidence that Senegalese-born Khaby Lame has the second highest count of TikTok followers globally at 138.3 million? He lives in Italy and makes videos of everyday problems, large and small. It is just these short clips, which hurt nobody, often verging on banality, that make TikTok successful. The fact that TikTok deletes certain terms related to issues that could displease Beijing, like those concerning the Uyghurs, causes neither protest nor outrage in Africa. The majority of TikTok users are between 18 and 24, and the short clips show lots of dancing, mime, and self-promotion. But there are already efforts to bring short news broadcasts to users on TikTok. TikTok poses both a personal risk of addiction and a geopolitical threat because Beijing’s faith in the internet is associated with an anti-liberal ideology, as Harvard lecturer Niall Ferguson warns. The competition here is between a censoring TikTok, controlled by Beijing and opposing free debate, and US-American Facebook, whose interest is in click counts and whose algorithms therefore favour violence and brutality. It seems young users cannot choose but between digital pest and cholera.
Why are there still no European search engines? Or a service that is as popular as Facebook? Instead, secretive companies from the US, the motherland of modern democracy, and the People’s Republic of China, a dictatorship, dominate. The attempts of individual governments, such as Ethiopia’s, to found their own social networks may meet with derision because they appear technically helpless. But they also appear to be the expression of deep insecurity concerning the fragility of social cohesion in societies that are already economically and ethnically polarised. Nick Clegg, the sleek former Deputy Prime Minister of the United Kingdom and now one of Facebook’s presidents, says, “I would love to see the next Google or Alibaba emerge in Europe.”
That sounds generous and pro-competition, but does not change the dominance Facebook exercises. Frenkel and Kang write, “Facebook had become as powerful as a nation-state, bigger than the most populous country in the world. But nations were governed by laws, and their leaders invested in public services like firefighters and police to protect their citizens. Zuckerberg wasn’t taking any responsibility to protect Facebook’s members.” Clegg, who is such an important lobbyist for Meta, then says soothingly, “Private companies really shouldn’t make such important, far-reaching decisions on admissibility of content by themselves.”
If Europe is about to lose Africa, it is also because it does not behave in the same casual manner as Silicon Valley and Beijing. Yet solutions from Europe could be more convincing than a Mark Zuckerberg or a Xi Jinping striving for dominance. It is a comfort to know that some of Africa’s elites agree. Unfortunately, there are fewer and fewer of them.
– translated from German –
Christoph Plate is Head of the Konrad-Adenauer-Stiftung’s Sub-Saharan Africa Media Programme based in Johannesburg, South Africa.
Choose PDF format for the full version of this article including references.