Taking on the Tech Titans
Reclaiming our data commons
Ben Hayes, Ben Tarnoff, Vahini Naidu, Anita Gurumurthy, Caroline Nevejan, Nanjira Sambuli
24 November 2020
Our webinar Taking on the Tech Titans: Reclaiming our data commons explored who owns our data and why it matters, the relevance of data extraction for countries in the Global South, and the impact of COVID-19. What strategies, structures and institutions are needed at national and international levels to confront Big Tech and advance digital justice?
Ben Hayes, TNI Associate, Founding director of AWO, a new data rights agency gave an introduction that addressed what makes the issues around big tech so important in the current climate. COVID-19 has crystallized the intersection between corporate power, state power, technology and rights. In the last decade, this area was characterized by Edward Snowden’s revelations in the early 2010s and the extent of government surveillance and the global state of surveillance infrastructure. In the last few years, we’ve gained a growing understanding of the power of big tech companies, the business model of surveillance capitalism and how it is affecting – or has the potential to affect – almost every element of social and material life in ways we don’t yet fully understand. COVID-19 has married state surveillance, corporate power and surveillance capitalism in stark ways. We have seen private companies like Palantir and Google shaping government responses to the pandemic, granting states access to data on our populations in extraordinary ways. Discussions about immunity passports and government access to data in COVID tracing apps are concerning the data rights community.
Ben Tarnoff, tech worker, writer and founding editor of technology magazine Logic, opened the discussion by offering his insights into how the present crisis is intensifying the power of big tech and how that intensified power will in turn have major consequences for data enclosure and extraction. Currently, a substantial share of humanity is living under some form of lockdown order. A clear consequence of this is a sharp increase in global internet usage anywhere a lockdown order has been implemented. When China locked down Hubei in January, mobile broadband speeds dropped by more than half due to network congestion. In Italy, home internet use has increased by 90 percent and global internet traffic is about 25 to 30 percent higher. Subsequently, we have also seen surging use of cloud services as the infrastructure that sustains the services people are using is taxed by the increase in users. As of late April, Synergy Research Group estimated that the cloud industry has already spent more money on data centers this year so far than in the whole of 2019. Two of the largest cloud providers, Amazon and Microsoft, have some 50 new large scale data centers under construction globally.
As well as increased internet use and demand for cloud services, we are seeing new demands for more invasive state and corporate surveillance. From contact tracing from mobile phone location data that has been particularly aggressive in Asian countries to monitoring body temperatures with thermal imaging and tracking the infected with facial recognition – something that has been explored in the US by Clearview AI. In addition, we can see greater surveillance of protesters and people engaging in any type of dissent or oppositional activity, particularly as uprisings are taking place around the world in the wake of George Floyd’s murder.
Having outlined the ways COVID-19 will accelerate digitization and give corporations and governments more opportunities to create more data about us and the environments we inhabit, Tarnoff went on to discuss why big tech are creating this data and what they are doing with it. When we talk about the tech titans and the possibility of reclaiming our digital commons we must understand how these commons are being produced in the first place and for what purpose. In order to do that successfully, we need to talk about capitalism since as long as capitalism has existed data has helped it grow. Bosses watching how workers work and rearranging them to be more efficient is a good example of how surveillance has generated information used to increase productivity. In the early 20th century, Frederick Winslow Taylor made systematic surveillance of the productive process a key part of what he called scientific management – a set of widely influential ideas about how to increase industrial efficiency that is still visible in global industrial processes today.
That data has been useful for capitalism is nothing new, however, the scale and significance of data has developed thanks to information technology. Digitization has made data more abundant since it has become much easier to create, store and transmit. A small device can be attached to almost anything to stream data from it including shipping containers, assembly lines, gas turbines and the wrists of factory or office workers. The ability to extract information from the productive process in order to optimize it has reached a new level of sophistication. However, observing the productive process is not the only way we create data. More broadly we create data whenever we do anything that is mediated or monitored by a computer. This encompasses a growing number of activities around the world. Even if you are not directly using a computer or you are not aware of it, data could still be created about you with or without your knowledge or consent.
No single technology contributes more powerfully to digitization than the internet because it contributes to the flow of data and creates more of it. Everything we do online leaves a trace which in part is why corporations are so eager to increase the amount of time we spend online. At one level when we talk about big data we mean literal bigness – more data, bigger data sets – but big data also means that the data can be made more meaningful and yield valuable lessons about how people or processes behave or how they are likely to behave in the future. This is true because we have more data, faster computers and developments in fields like machine learning have given us better tools for analysis. Machine learning in particular is a powerful tool for pattern recognition that can be trained on data. The system learns the pattern and can then recognize that pattern in the future, with faces for instance.
The bottom line is that big data is driving digitization because any piece of information when combined with other information and analyzed or interpreted en masse may reveal actionable knowledge about the world. This information might teach a manufacturer how to make a factory more efficient, an advertizer what kind of things you might buy, or enable an intelligence analyst to map your movements or a drone operator decide where to drop a bomb. For these reasons digitization is becoming as important to capitalism as financialization was during and after the 1970s. As well as offering a new engine of capital accumulation, it gives states new tools for social control to help order populations to be more useful for the purposes of accumulation and manage redundant, rebellious and racialized populations.
Vahini Naidu, trade negotiator, Department of Trade, Industry and Competition in South Africa moved the discussion on to how big tech, largely concentrated in the North is seeking to consolidate its power at the expense of development in the global south.
Contrary to popular belief, the World Trade Organization (WTO) is still active. Despite one major developed economy actively seeking to dismantle the Appellate Body, the enforcement arm of the WTO, they are also aggressively advocating for digital trade rules in a trade agreement. The development agenda on eliminating agricultural subsidies and intellectual property rights is under discussion and the pandemic has reinforced the importance of this agenda for developing countries especially from a food security and industrial policy point of view. Advanced economies are also pushing their own reform agenda and negotiations in the WTO by aggressively pursuing rulemaking on e-commerce investment transparency and trying to eliminate any special preferences for developing and least developed countries. These are all highly contentious issues that the membership does not agree on.
The relevance of the WTO has also been reinforced by the rise in unilateral measures adopted by some advanced economies while developing countries do not have the legal capacity to justify their policies. Since July 2016, the US has tabled a submission attempting to refocus the existing exploratory work program on e-commerce to begin to examine the linkages between digital trade and economic development. This submission has listed 16 seemingly innocuous trade-related policies they wanted countries to adopt. A series of proposals by advanced economies soon emerged, branded as e-commerce for development despite not including any rules on development.
Since 2018, a group of members have decided amongst themselves to negotiate multilateral rules on e-commerce and currently, 84 out of the 164 members in the WTO have become formal signatories to these plurilateral negotiations on ecommerce. Critical to point out is that there has been a large influence from the international chamber of commerce, big tech firms and corporations involved in advancing their interest in the negotiations. A common theme in these negotiations has been to pursue a deregulatory approach to digital trade. The current text strongly favours rules that leave how they use, share, treat and commercialize technology and at their discretion, with a minimal role for state intervention. In the currently proposed text, there are very limited exceptions that have proven through WTO jurisprudence to be extremely difficult to prove, especially for developing countries.
Naidu emphasised that even though these negotiations are branded as e-commerce they also touch on internet governance, ICT matters, products, services, technology and digital trade. Big tech have demanded rules including not imposing any customs duties fees or charges on a digital product transmitted electronically and leaving electronic authentication to companies and consumers to decide how secure the electronic transactions should be. As we know consumers are not in a position to ensure this. Other rules include things like no member shall restrict or prohibit the cross-border transfer of information including personal information by electronic means if this activity is for the conduct of a business. Members are required to ensure cross-border data flows to facilitate trade in the digital economy and no member shall require the transfer of or access to source code software or to an algorithm expressed in that source code.
These are clearly extremely broad proposals and one must ask whether the WTO is the right forum to be discussing these issues. Several WTO members including South Africa and India have not joined this agreement and expressed extreme concern with how these negotiations have been conducted. South Africa believes that these multilateral rules on e-commerce and digital trade are premature. We have seen several challenges emerge from the current reading of the rules, the first that there are a range of unanswered technical questions relating to the ownership control management of data that will be critical for evidence-based policymaking. Other questions raised include if the gains outstrip the costs, if the net value generated is captured in the domestic economy or flowing out, and if there are employment gains or not.
There are a number of policy and technical questions still being analyzed from a South African perspective. The content of many of these current proposals for rules in this area also suggests a trajectory opposed to inclusive and equitable economic development. We are seeing rules that would permanently curtail governments ability to manage digital transformation by imposing a range of prohibitions across the board on customs duties, local content, localization, transfer, data flow and source code disclosure. These rules appear to be designed to lock-in the advantages currently enjoyed by corporations and countries that already lead the digital economy and suggest that many of the proposals aimed at facilitating digital trade would simply reinforce existing imbalances.
Naidu pointed out that many of these rules will constrain developing countries’ abilities to build digital capacities as we see a huge asymmetry in the development of e-commerce globally. Two countries account for 90 percent of the market value of 70 of the largest platforms [LINK to list for more info]. These problematic asymmetries and the digital divide will further reinforce global social divides as well as determine the countries that will be able to recover from the pandemic. Data is at the heart of this and we know that digital capacities are needed to build productive capacities, especially for developing countries, for sustainable and inclusive growth. The push by big tech to liberalize data needs to be scrutinized and there must be more focus on what is happening in the WTO for developing countries to begin to develop their own frameworks.
Anita Gurumurthy, founding member and director of IT for Change, in India went on to bridge the background that Ben Tarnoff laid out and the specifics that Vahini Naidu highlighted.
There are two key aspects of current capitalist production that are different from ten years ago. The first is that data has become a valuable resource and AI based on data is an important factor of production today. We have an entirely new way of economic organization and a new value chain stemming from how production and market exchange are managed through insights from data. There is a huge industry for annotating data that is being collected, sorted, arranged and processed continuously. This is the industry for intelligent production that is defining what we call the digital economy. It is no longer about one sector in the economy, data business is cross-sectoral. One example is companies buying stakes in others based on a long data game to seize the market, taking a broad view of how to expand their dominions based on data marriages. This is clearly evident since Bayer acquired Monsanto to marry pharmaceutical data with soil and seed data to control more than a third of the agricultural input market or Amazon acquiring Wholefoods to marry its online commerce with a booming offline market for organic food. Data has become key to modeling the entire supply chain from what is produced to how, where and when you will sell those products. Data-based intelligence determines productivity in the digital economy and this is the new capitalist model based on data value chains.
The second aspect of current capitalist production concerns who controls these new supply chains. As previously mentioned, platform companies are the leading firms controlling data value chains, companies that used the internet to build their initial business and soon discovered the power of data to develop the presence to model not just their own business processes but the market itself. Platform firms dominate the global business landscape today capturing nearly 60 percent of market capitalization and their power is growing at breathtaking speed. As Jack Marr said in an interview is took the Alibaba empire 15 years to achieve the numbers that Walmart achieved in 60.
The belief that the biggest problem is the capture of our personal data and the control of our behaviour through its enclosure only part of the story. Information about us and the world is combined into raw material for capitalism to create and expropriate value for itself. The deep invasion of our lives is problematic but how such invasion is translated into exploitation and injustices needs to be fully understood. The first step to confront the fictions and fake news that concern data.
The first incorrect assumption is that data is non-rivalrous and the supply unlimited. In reality, data as a unitary piece of information is meaningless, it acquires power only when it is relational or is seen in relation to other data. This is why data is a system resource enabling connections to be made at different depths and layers within a system. The big tech companies founded years ago were able to build these systems of connections thanks to the network effects of the internet enabling them to keep data locked up and unavailable to others. The data from which meaning and value can be generated not only for economic but also social and public purposes does not come with any authorship or creator rights. Big tech can simply mine it for profit. With this absence of property rights, we are seeing the de facto enclosure of data.
The second incorrect assumption concerns data flow. We are told that the data economy presents great opportunities through free data flows across borders – Bangladeshi data going to the US as easily as vice versa – and the internet has enabled a brand new frictionless economy. However, the data economy, very much like the knowledge economy, is far from frictionless when we consider how the early movers like Google and Facebook were able to capitalize on network advantage and the ability to expand markets at near-zero costs on the internet. Their platform infrastructures became the apparatus par excellence to gather and hold data and build data-driven economies of scale. It is hardly possible for a Bangladeshi firm today to build even a small unicorn startup that can attract venture capital funding, let alone another Google. In the intelligence economy where data insights are monopolized by a few firms, market structures grow to become unjust. 75 percent of the cloud computing market, 75 percent of patents and blockchain and 50 percent of spending on the internet of things is shared by the US and China. If data did not flow freely the grand economic rents that the big firms are getting would simply stop which is why the major powers are pushing the fiction that data cannot be claimed as a resource that belongs to people and their jurisdictions.
Caroline Nevejan, Chief Science Officer at the City of Amsterdam and Chair of Designing Urban Experience at the University of Amsterdam was next to offer her perspective focusing on the role of big tech in the city and some of the challenges cities like Amsterdam face from the power of big tech and different models of using data to support democracy and citizenship control.
The fact that we have this new media does not just mean bad news, in principle it allows us to gain a new understanding of the world. For a long time, humans were the measure of everything. We were then able to measure blood pressure, distance and among other things but now because of data and the internet, we have a measurable humankind. We can know what many people feel, think and see, a potential we must pay attention to.
However, in our cities especially we have now moved into complex systems that mean we now live in communities where such systems have more agency than the people who live in the city. This raises the question of trust and how to design for participation and surveillance. In this new meta-design paradigm, every step you take on the way becomes a new form of participation or a form of surveillance, something we are struggling to know how to talk about.
Regarding data and big companies, as mentioned the financial economy used to make up a third and two-thirds was the sharing economy. In the last decades, big tech companies have invaded private lives as well as the sharing economy, financializing many elements of our lives that previously had nothing to do with money.
What we see simultaneously in the city is people’s data taken by big tech companies as well as payments with bitcoin for houses resulting in a whole market that the city is not involved in. The city is a place dependent on taxes to build schools, hospitals and roads so just as our inhabitants need to be here, so does their data. We are our data and by taking personal data as well as introducing an uncontrollable financial system that does not allow for tax-paying cities are becoming victims of big tech greed.
We see too much data from the privacy sector and as we lose our privacy we cannot doubt or change our opinion and thus democracy is at stake. Concurrently, we have too little data about what’s happening within companies and thus justice is also at stake.
The question is how do individuals deal with this. We see global efforts to invest in new architecture including creating digital commons, making new software, the Linux community and the open-source and blockchain communities. People are also protecting themselves with crypto to try to make their own lives outside the control of big tech. The Centre of Investigative Journalism offers excellent courses with strategies on how to do this. It’s vital to teach and develop new strategies to reveal hidden truths and create space where there is a limit to how far big tech can invade.
What we see is a tension between cryptography and the public domain. However, important legal instruments like the General Data Protection Regulation (GDPR) in the EU are changing how the data market, companies, cities and governments behave. The Right to Information Act (RTI) has also enabled positive changes in many countries but is now jeopardized because of digital technologies are not designed for transparency.
Where do we go from here and how can we work to find new ways to use this data potential? The city of Amsterdam is looking at how rhythm can be a way for people living in the city to understand their lives better, tuning the rhythms of nature, social life and culture.
We must realize the issues are as much about our environment as data and work on the idea of multispecies urbanism, that first considers the ecological design before how people will integrate into it, let alone technology. Attention must be given to looking at the environment around us in new ways.
This principle also applies to how we deal with each other. Big tech designs for the same kind of people but it is fundamental that we design for diversity in order to thrive.
The municipality of Amsterdam has taken the decision to become a circular city, using the model of Kate Raworth’s Doughnut economics enabling a relation between social and ecological as well as local and global. In this way, we can start reflecting on how we operate as our own footprint develops.
Nevejan concluded her intervention by highlighting the need for a new awareness of climate change. Applying imagination to our ability to perceive humankind and our actions must become a tool for our survival.
Nanjira Sambuli, researcher, writer, policy analyst and advocacy strategist on tech and governance shared her reflections and perspective on how international civil society can respond to the assaults of big tech.
First and foremost we must consider and accommodate the diversity of perspectives on data and privacy in the digital economy, situating them in the historical, political and nuanced contexts of digital developments across the Global South and among the diverse and marginalized minorities around the world. It is essential to reflect on how we study their understanding of these issues and how we measure our progress and success. We can advocate for laws and regulations protecting data and privacy and consider them successful when we see countries putting them in place but they may not be reflective or representative of how citizens from the Global South perceive these issues. The actions and recommendations that we put forth in countering or reforming the status quo can sometimes be impractical and even come across and callous for most developing countries. A good example of this is the deleting Facebook movement that became a moral imperative but is impractical for many people for whom Facebook and its family of apps is the entire scope of the internet. Studying the use of WhatsApp in Brazil has shown that removing it from favelas amounts to removing a resource people use to shape their realities. The alternatives we propose to reclaim our data commons or the digital space need contextualization and humility in how we put them forth to create room for the discussions of those who cannot adopt what we mainstream. For many people the entire internet experience in an enclosure altogether.
Developing markets are the greatest growth potential or capture trajectory for the digital and data economy and the mentioned assaults are coming from the private sector and Western and Eastern governments turning the Global South into the next geopolitical battleground. If we reflect on the relationship being set up, we are not even users anymore but slaves to be mined for data without receiving anything in return.
In this configuration, civil society must not repeat the same mistakes we have seen before by speaking or advocating for those left behind but rather create room to amplify the voices of those observing and mobilizing on the ground. People in the Global South and marginalized communities have been following how these issues are developing in their contexts but this is not often reflected in what becomes the internationalized movement. We see IBM stating that they will stop developing facial recognition technology while a group of AI researchers who happen to be women of colour have been speaking about this for a long time. They may not have gotten the deals to write the books we cite today about these issues but civil society has a moral imperative to bypass that. When we speak of principles and values like diversity and solidarity as a currency not everybody is using we must be careful not to seem as though we are also using a selective application and enforcement as international civil society. The marginalized majority should not have to fight the onslaught from the private sector and governments as well as internationalized or institutionalized civil society. Reflecting on and accepting this uncomfortable reality is where we must start to rebuild.
Ben Tarnoff was the first to respond to a question from the audience and Ben Hayes on how we can deal with the ownership of data if we do not also address the ownership of tech infrastructure.
Thinking about infrastructure as well as data is an important point that connects to the imperfect metaphors that we use to talk about the process. Data extraction, data commons and data enclosure are metaphors that suggest there is something out there called data that is extracted like oil or enclosed like land. The reality is that data is actively constructed beginning with a camera that observes us or an algorithm that tracks our mouse movements around a website. These observations reflect a very intentional and purposeful process. In the next stage of the pipeline is what in machine learning is often called data wrangling, turning raw data into the kind that is useful for analytics. Data scientists spend most of their time on this process known as feature engineering that isolates and extracts the particular attributes of the data that will have maximal value for whatever is attempting to be predicted. We must remember that data is not something preexisting to be collected by corporate and state actors that we can take back but is completely intertwined with infrastructures that create, organize and draw value from it. The question is not simply to socialize the data commons or assert our ownership over existing data. We need to move upstream and interrogate and reorganize the structures that are creating that data in the first place.
Vahini Naidu continued by offering her perspective on the possibilities of digital sovereignty and the dangers of the WTO plurilateral discussions on e-commerce. Before outlining some of these problems, Naidu encouraged webinar participants to follow her previously provided lists of links to gain a deeper analytical and policy perspective on why some of the proposed rules would be extremely problematic for developing countries. The pandemic has revealed the deep structural vulnerabilities in the global economy seen even before the move towards consuming locally and producing within a nation’s borders on account of environmental considerations amongst other things. However, it is also an opportunity to look at how developing countries can grow their domestic industries especially in terms of the digital economy. No developing countries are even close to achieving the mass scale digital industrialization success that has emerged from the deliberate state interventions by the Chinese government.
In terms of data, the biggest issues now in the WTO is whether members will agree to unfettered cross border data flows. It is extremely premature to impose these kinds of rules when so much remains unknown. The international engagement should for now focus on experience sharing analysis, dialogue and possibly also cooperating in some areas. An additional problem with WTO rules concerns developing countries abilities to retain the space to develop and implement policies with flexibility to change, tweak and in some cases even abolish them. The levels of uptake of legislation in cybersecurity, electronic transactions, in data legislation must be taken into consideration. For example, only 50 percent of African countries have legislation on privacy and data protection. 72 percent of African countries have cybercrime legislation but implementation levels are very low.
More needs to be done in terms of policymaking to recognize the sovereignty of national data. The localization of data is crucial because even if data is located outside national boundaries and a country can access it, others for example those that own the cloud also have access to it. In the context of a growing digital and technological divide, these firms will have a much higher capacity to use and profit from that data.
From the revenue perspective, we are seeing the EU take the lead in taxing the digital activities of big tech and we should also look at how these activities can contribute to the tax space of developing countries including a customs duties moratorium on e-commerce.
In addition to responding to Ben Tarnoff’s thoughts on infrastructure, Anita Gurumurthy shared her final view on a new data settlement that simultaneously enhances our data rights while pushing back and reigning in the power of big tech. The good news is that data concentration is being met with increasing disapproval in the past few years and the GDPR had the idea of self-determination and sovereignty limiting the kind of data that could be collected by platforms. However, this has not mended the platform economy because informed consent is a blunt instrument essentially amounting to nothing. Other measures are encouraging including Elizabeth Warren’s proposal to designate certain platform infrastructures as essential utility and legal scholars in the US recommending a structural separation of platforms and commerce that would ban companies from operating a platform and marketing their own goods and services on it. Duel services tax has been introduced in ten countries and French antitrust regulators fought for Google to remunerate press publishers. It is becoming increasingly clear in dealing with big tech power that not just market failures are at stake but widespread economic catastrophe.
Small countries will be browbeaten to sign up for new e-commerce trade deals and these instruments will not take care of the incentive structures of big tech. They will simply pay the tax and write it off as a cost to the company and get on with the business of data theft and its various positive externalities for colonizing the planet. IT For Change has been looking at the need to create a distance between each of the major functions or layers of the data economy and start to take back from the colonizers who have already built profitable empires with the raw material from our soil. The data collector or holder cannot also build and operate cloud infrastructure and provide AI services. These proposals and regulations can help but to create just and fair economies we need to free up data power and intellectual capital for economic, social and public value all over the world.
In February, the EU came up with a plan called the Data Strategy about various ways by which to reclaim data power. IT For Change would like to propose that in addition to principles of sovereignty countries also complement their regulatory and legislative framework with economic rights about data. This is crucial because sovereignty must be combined with economic power and defined community ownership rights in data. These rights can be about access that a community has to data, the benefit extraction possibilities through use and licensing, defensive rights and boundary management. There will always be conflicting rights, individual sovereignty, contractual and public interest claims. An independent data governance authority and data trusts are needed to manage these rights through nested sovereignty principles that deal with the data economy not only the personal data rights.
Caroline Nevejan offered her final comment by elaborating on a few ways Amsterdam is using or planning to use data and emphasizing the importance of the locality of data and data for transparency and against corruption. In the Netherlands, there are big issues with tax avoidance made more difficult to tackle due to data sources and sharing information. However, a positive step has been a position posed by the deputy mayor and accepted by the council to enforce signs at the entrance of streets warning people they will be filmed as well as the development of the GDPR, that although is not good enough is at least a first barrier. We lack the arts in the data scene and need good visualizations that empower people instead of baffling them with schemes that are hard to understand. It is also very important to be aware of who has access to what data and since, as previously mentioned, WhatsApp and mobile phones are the entirety of the internet for most people we must make sure that WhatsApp is not used to arrest young people who protest for example. Bringing the tech titans under democratic law is fundamental to force them to accept responsibility and accountability. There must be a right to address and a right to personal sovereignty in all the interdependencies we have.
Nanjira Sambuli concluded the webinar with her final reflections and by addressing a question from the audience on how can we hold big tech to account for its role in promoting violence internationally. We do need to acknowledge that it is clear that there is discontent and people are showing various ways to push back against this onslaught coming from big tech. There are many conventions and international platforms that need to be turned into action to fight this marriage between the systems of governance and corporations calling for the rules to work in their favour. Legal systems must work for the citizens as they were initially envisioned to. The pandemic has forced protesting citizens from the streets to the platforms where these corporations are making more money in the name of our protest. This circle complicates the equation but we must start reclaiming government and governance spaces to truly use them to support and represent citizens. Prescribing strategies is difficult as there is no one size fits all model and a varied toolkit is needed for different countries. Keeping hope alive and getting strategic and creative is how we must move forward.