17 November 2020

Is Big Tech Setting Africa Back?

By Nima Elmi

Technology transcends national borders, so it shouldn’t be surprising that, in 2020, private technology and telecommunication companies control more data on the average person than governments do. And it’s not just data ownership: Policymakers similarly fall behind in understanding the power of data. That’s a problem, particularly as social media platforms are able to influence political outcomes with few to no repercussions and geopolitics becomes a duopoly of technological trailblazers like the United States and China, with the rest of the world looking on as spectators—and reluctant participants.

The coronavirus pandemic has moved many people’s lives online and demonstrated the revolutionary power of technology in driving economic growth despite physical stasis—whether in e-commerce, continuing employment for those able to work digitally, or virtual schooling. This dramatic shift wouldn’t have been possible without artificial intelligence (AI) embedded within now-essential services like Alexa, Siri, and Zoom.

Artificial intelligence is usually talked about in sensationalist terms. But hyperbolic language can mask simple business truths: AI capabilities—understanding everything from shopping habits to future careers or propensity for criminality—will only ever be as good as the datasets that feed them and, without diverse data sets, the ability to innovate and enhance existing AI functionalities is limited. The de facto U.S.-China AI duopoly doesn’t accurately represent the cross-cutting, global consumer bases tech companies serve; for the game to go on, spectators need to pitch in—lest they lose their pastime.

It is against this backdrop—a thirst for new data to keep the AI engine chugging—that colonialism has morphed into its latest form: data colonialism. Data colonialism is driven by the control of data as a proxy for the control of people and is quickly becoming the reality faced by many emerging economies today. Far from decolonized, these countries are subject to the whims of Big Tech’s unfettered rise; vulnerable peoples’ data is used to enhance companies’ innovations, entrench their economic and political might, and, in effect, occupy the daily lives of billions of people. Data extraction, monopolization, and monetization are data colonialism’s core tenets.

Data extraction, monopolization, and monetization are data colonialism’s core tenets.

Africa is ground zero for data colonialism. It is the continent with the largest number of countries; most cultural, linguistic and racial diversity; least connected nations; and its data protection regulations range from limited to nonexistent. Africa has always been a continent rich in natural resources, and, today, the diversity of the continent’s population renders it equally rich in data resources. But Big Tech’s exploitation of this diversity—heralded under the guise of internet-for-all initiatives—actually undermines the data sovereignty of African nations and impedes their ability to develop their own digital economies. That’s hardly a tide that lifts all boats.

In 2017, data overtook oil to become the world’s most valuable commodity, but it remains hard to define. At its most basic level, data is us—humans—in digital form: how we look; the languages we speak; our viewing preferences, medical records, voices, learning habits, and music preferences. For AI, all data is valuable fodder—enhancing algorithms’ capabilities and generating revenue for the companies that collect it.

When it comes to obtaining this data, the latest figures from the U.N. Conference on Trade and Development highlight just how vulnerable Africa is in comparison to its European counterparts. In Europe, 96 percent of countries currently have data protection laws in place—versus only 50 percent in Africa. Globally, only 43 percent of the U.N.’s designated Least Developed Countries (LDCs) have such laws; 33 out of the 47 LDCs are in Africa.

As well as the unrivaled, diverse datasets that Africa offers tech companies, the continent’s relative absence of data protection policies—and a limited understanding of how valuable and influential data can be—is a central causal factor behind its vulnerability to data colonialism. There are myriad reasons for this vacuum in regulation on the continent, including a general lack of capacity, conflicting priorities, and political instability.

Data colonialism is at times conflated with humanitarian efforts—after all, who can object to expanding digital access, which nearly all agree is crucial for economic growth? But corporations’ support for emerging economies may be a double-edged sword, as data extraction undermines African nations’ abilities to develop indigenous digital economies that can enhance their own capabilities.

It’s farcical to assume corporations’ forays in Africa are chiefly compelled by goodwill: Corporations are ultimately accountable to their shareholders and driven by their bottom lines. That comes at a cost, of which Africa bears the brunt.

Both Chinese and U.S. tech giants adopt a humanitarian framing in marketing digital infrastructure projects, but approach the ulterior motive of data extraction differently. Chinese corporations leverage political initiatives like Beijing’s controversial Belt and Road Initiative to secure contracts with African governments, whereby AI and other technology solutions are provided in exchange for access to local citizens’ data—like the deal between CloudWalk, a Chinese developer of facial recognition software, and the Zimbabwean government. Firms are relatively transparent about their goals, with little to no checks or accountability on how the information they receive will be used.

Western corporations, by contrast, adopt a more covert approach, establishing digital hubs and offering free internet access. These include Google’s Internet Balloons and Facebook’s Undersea Cable, which encourage more Africans to get online, use their services, and, by convenient coincidence, relinquish their data in the process.

Data colonialism hinders Africans’ abilities to develop their own technological innovations based on their indigenous datasets, which have already been exploited by the more established hegemons of the tech world. Moreover, it relegates African nations to forever being consumers of foreign tech innovations that are developed using their own data and then sold back to them. This is akin to traditional colonialist practices that seized raw materials from the African continent, refined and manufactured them into goods in the West, and sold these goods back to the same nations from which they were stolen—leaving them unable to develop their own manufacturing capabilities. They became eternal consumers with underdeveloped economies as a result.

Data colonialism hinders Africans’ abilities to develop their own technological innovations based on their indigenous datasets.

Beyond ostensible humanitarian efforts, tech giants’ expansion of digital infrastructure also serves to skirt and undercut regulatory activity. That’s because, while there is little to no data protection or privacy regulations in Africa, nearly every African nation has a telecommunications regulator. In some ways, this is understandable: The penetration rate of mobile phones on the continent is 77 percent, while only 26 percent of Africans use the internet.

While telecommunications companies are able to collect data on a far greater number of Africans than are tech companies, the depth or value of the data they collect pales in comparison to that mined by technology platforms. This is because mobile phone operators have access to data that is comparatively limited: subscribers’ identities, addresses and banking details, cell site or geolocation, call logs, and other technical data. Tech platforms, by contrast, can access more granular information about their users—from political views, relationship status, friendship network, and travel preferences, to mental and physical health characteristics. Consequently, while telecommunications companies can use their data to inform basic research and development efforts, tech platforms’ data not only informs what types of products to develop, but also how to market those products directly to consumers.

In fact, telecommunications companies are often held to ransom by their regulators—with threats of license revocation or administrative delay should they fail to meet regulators’ demands—from paying higher licensing fees to rolling out infrastructure in rural areas. Technology platforms, by contrast, have little to no barriers to entry—and also no regulator to hold them accountable. For this reason, tech platforms seek to minimize their dependence on telecommunications companies by developing initiatives that provide internet access directly to end-users without a need for mobile subscription. After all, if people can connect directly to the internet through a Balloon or fiber optic connection, the need for mobile phone subscriptions will decrease—particularly given the ubiquity of services such like WhatsApp, which has over 2 billion monthly users.

When it comes to empowering indigenous digital economies—or not—chronic avoidance of telecommunications doesn’t help. Telecommunications companies are often registered corporate entities in the jurisdictions in which they operate—contributing to taxes, employing local staff, and contributing expertise and capacity-building. The same can’t be said for technology companies in Africa—which often import their own staff from overseas for any local activities or fail to set-up a local entity to provide their services.

One way to equalize this would be to regulate Africa’s tech sector by holding tech companies to the same standards as telecommunications operators. African governments need to improve their understanding of data and its power—and educate the public about the advantages and pitfalls that come along with it.

African governments need to improve their understanding of data and its power—and educate the public about the advantages and pitfalls that come along with it.Much like within the European Union, equal efforts should be given to developing national, regional, and pan-African data protection and privacy laws.

Some governments are already trying to fight data colonialism through regulation. Nigeria, Rwanda, Kenya, and South Africa have advocated for data localization, which requires all data collected on their citizens and in their countries be kept on servers where they are collected—rather than far away in California or Guangdong. This is a good and a forceful first step, but cross-border data flows are crucial to the modern economy—and data localization laws are generally short-sighted tools that can stifle, rather than foster, indigenous digital ecosystems.

Instead, African policymakers should develop a pan-African approach to cross-border data flows—possibly within the African Union—by understanding that data is a commodity worthy of protection at national and regional levels, and perhaps even under the African Continental Free Trade Agreement.

African governments do not need to replicate the European Union’s General Data Protection Regulation, but they should design and implement an indigenous data protection framework that suits the continent’s context and capabilities through public-private cooperation.

More than anything, Africa requires transparency. If leaders don’t act swiftly and proactively to address data colonialism, Big Tech could compound the effects of traditional colonialism and limit Africa’s ability to harness technology on its own—forcing the continent to lag behind the rest of the world forever.

No comments: