Faced with a rapid, partially invisible pandemic, a majority of countries have so far chosen emergency measures based on territorial zoning, which does not discriminate between the sick and the healthy. Now, digital contact tracing is being presented to us as the ideal way to differentiate between these two groups of people. In reality, it instead adds another political problem, that of an enlarged surveillance apparatus, without solving the underlying one: the lack of an effective public health strategy.
The Covid-19 pandemic has profoundly changed our relationship with life and the horizon of death. Daily life and the banal gestures of globalised capitalism have been changed practically and are no longer discussed only intellectually. Moreover, they have been changed in a hurry. The response of many governments to the pandemic has been severe containment. The free movement of people has been curtailed in many places by emergency legislation.
The choice of containment — due to the lack of vaccines and the lack of strategic reserves of available tests and hospital beds — has certainly slowed the spread of the virus. However, it has also added the problem of stopping many economic activities that depend on free movement. One might have thought that an intense debate on the way out of the crisis was going to begin, and in some regards it already has. But at the same time circles of power seem to have consolidated their fascination with the technological solutionism of computer security.
Those in power seem to think that the revolution of artificial intelligence, big data, and traceability will save us, or more importantly will save the status quo by favouring the same companies already in place. But how can anyone believe, and make others believe, that the digital tracking and detection of potentially-infected individuals will make it possible to re-establish a “normalised” health and economic order, with an acceptable victim rate? Or that this tracking means we can soften the measures without too high a risk of resurgence?
It’s doubtful if a state of “normality” would even be possible through a surveillance network of tracing applications. Would we really be able to regain our freedom of movement if that movement is under constant surveillance, governed by digital applications? As many security professionals currently seek to reimagine our future, will we let them treat us like herds in a pasture by coupling each person’s biological identifiers with their digital identification?
In this article I will argue that we should not accept this kind of despotism, fuelled by digital technology, but that instead we shall respond to the pandemic by putting health and digital tools at the service of citizens. Doing so would involve following the lead of French sociologist Dominique Boullier and his work on citizen-centred design.
Containment variations a result of poor planning?
Due to the scale of the coronavirus pandemic — which far exceeds previous pandemics — many states have been unprepared, leading them to fall back on the most traditional and authoritarian containment measures, confining entire populations in their homes and severely restricting their freedom of movement. This was not so much an optimal strategy but rather the only option available to them due to the lack of effective testing, high-performance masks, and hospital equipment for respiratory emergencies.
In contrast, the governments of states which had tests, masks, and hospital equipment ready when the virus hit, opted for a very close follow-up of infected individuals and their recent contacts. They used a reasoning based on the networked spread of the pandemic and not on geographical blockades, which only temporarily slow down the progression of the virus. Physical distancing was used and enforced, but freedom of movement was higher than under a blockade.
In Europe, countries like Germany, Sweden, and the Netherlands benefited from policy decisions of the past which, by comparison to other countries like France, had put less emphasis on austerity “cures” such as reducing the number of hospital beds. They were thereby able to avoid complete confinement orders.
Unity in the folly of technological solutionism
Despite the diversity of situations in Europe — we have clearly seen structural differences in terms of public health and economic robustness — a kind of consensus has emerged. It revolves around a form of “technological solutionism” and presents digital tracing applications as a way to overcome all the problems this pandemic has tossed at us. In his book on the moral consequences of digital technologies, Morozov already brought to light this shared “folly” of technological solutionism. This is the belief that the digital supersedes the physical and that you can solve every problem with a click.
In the field of security, ranging from sanitary emergencies to defence and police, a similar belief has emerged. I have previously shown this to be a product of a transnational guild of digital technology professionals involved in security matters and of their influence over power circles to manage border controls, security, and interoperability. As this group — which includes strategic communications and advertising companies — is involved at the highest level of government in different EU countries, its influence on managing the crisis ends up superseding the influence of health professionals. In the absence of a vaccine, they turned social acceptance of the confinement measures (and their rebranding of it) into something radically new. The priority communicated in the media seemed to be that the positive news of a digital solution should make up for the bad news of high death tolls, and so the countries worst hit by the pandemic were most keen on counterbalancing their poor performance with a brand-new and ever-so promising app.
A collective belief that the internet, algorithms, and artificial intelligence will beat the virus
Hundreds of proposals came from all over the world. The computer industry — from the most powerful players (GAFAM in the lead, but also industries involved in cyber defence) to young “start-ups” — is now promising to be able to track individuals. They propose to do so by identifying and geolocalising them, tracking the actions of their data, reconstructing their past activities on the net and elsewhere, and especially by associating these data with health profiles. This might suggest whether these individuals were, or are, sick (declared or not) and if they are in low- or high-risk areas as the virus spreads, even if in such models there is hardly any area free of risk.
Following the spread of the virus will therefore be based on data from multiple web servers, centralised or not, through the use of mapping tools and linking the status of a specific person with their travel via a specific application — either with a smartphone or an autonomous Bluetooth card, as considered in New Zealand. Some applications envisage to also have real-time “flags” about the status of individuals who are obliged to move and have contact with the public. While some countries are (considering) making participation mandatory, most democracies are pursuing voluntary applications based on “privacy by design” to respect the GDPR and more generally the privacy rights of its citizens. But the fact that they therefore choose to anonymise the data collected is not proof that privacy will not be affected or that discrimination will not occur.
Notice how quickly policymakers have officially accepted and launched media campaigns that the internet, algorithms, and artificial intelligence can block the advance of the virus. Maybe they think that even if the selected apps will not really work in practice, they might nevertheless assuage the critics of severe confinement and create some hope in the public that it will recover (parts of) its mobility.
Costs of compliance and other hidden motives
Unfortunately, this mobility is not a sign of freedom of movement enjoyed as a right. Rather it comes with the cost of compliance, including the cost of providing personal data and accepting surveillance measures for the greater good. Tracing applications do not contribute much to public health, however, if they are not combined with widespread testing. Instead, they can also be seen as a building block in the management of public opinion, as if participating in the (voluntary) application was de facto a poll in favour of the government and an acceptance of its emergency powers.
A ‘bio-digital-surveillance’ has emerged from the Covid-19 outbreak as the only political ‘option’ ostensibly suited to relax confinement, even in democratic settings. But this political thinking has simultaneously created confusion about the functions of the digital apps and the medical strategies they are supposed to implement. Their purposes are multifold and not always oriented towards health. Some of these surveillance tools are tracing and localising sick individuals, while others are preventing them from moving without warning. Some are also designed beyond the scope of individual movement to trace personal encounters, with the possibility of demanding that a visit be paid to the hospital, police station, or town hall to register as potentially sick. Some of the apps also intend to identify via algorithms all possible unexpected encounters with a sick person (still without symptoms) and to transform these different encounters into cases of “suspicion”, without the cooperation of the individuals.
This digital tracing is presented as a great innovation, but in fact it is everything but new. Its digitality certainly conveys the image of instantaneity and efficiency, but the logic at work is old. Since the cholera epidemics of the 19th century, contact tracing has been considered a better, more targeted, and therefore more democratic practice than generalised confinement. It emerged from the production of specific knowledge about virus ecology and replaced the types of plague confinements we had before. In 1854, John Snow — one of the founders of modern epidemiology — established a cartography to validate his hypothesis of the role of water in the spread of cholera. But to track the virus, yesterday like today, it is necessary to collect much more data on the individuals who are infected.
What drives this grand digital strategy?
The key factor in discerning between the blind confinement of all people in a given area, be they sick or not, and the selective confinement of sick individuals is the visibility of transmission and its mechanisms. This is what eliminates false correlations and enables areas to “reopen”, at least for healthy individuals.
Data collection, mapping, and population tracing are only meaningful if they are followed by a systematic screening through tests that are applied to everyone and periodically renewed. Without this two-step strategy, of which the second is crucial, there is no point to develop the first.
As I have pointed out in my book “Data Politics”, raw data do not exist. Instead, politicians construct how to collect and organise data and the data in turn construct policies on knowledge foundations that are being built up and based on differentiated premises (data politics).
This is true also for health. Tracing apps do not directly oppose the free movement of individuals and are not based on their absolute confinement; they aim to create personal security belts around each circulating individual to avoid contamination. But they are a question of temporality as much as of space.
If there is one characteristic of modern pandemics, it is that in the absence of a vaccine, it is now instead possible — thanks to the speed of digital technology and data networking — to know more or less in real time the “channels” of viral propagation. These can be correlated with modes of transmission, identifying risk populations in real time and thus avoiding bureaucratic slowness (as Google Flu did, but unsuccessfully so). In theory, this enables authorities to anticipate the spread so that medical infrastructure is there when it’s needed, or at least faster than before (although, in the case of the Ebola epidemic, big data was actually of no help at all).
It is therefore not surprising that all states with tech industry capabilities are seeking to combine digital traceability with the imperatives of non-contamination in order to restore a degree of conditional freedom of movement. The freedom to work is deemed to be of particular importance in that regard, in order to limit the economic effects of containment.
It can also be argued that the focus on the digital is all the more emphasised since the resilience of the hospital sector has previously been so weakened by austerity policies (a card France and the UK seem to be playing). The focus on the digital becomes a “mask” for past failures and risks being a pure simulation policy if digital tracing is not linked to a generalised testing policy. This is because digital data are not capable of detecting anything if they are not accompanied by concrete and widespread measures, such as the test booths used in South Korea.
Urgent need for a citizen design protocol to prevent a control society
As Dominique Boullier has forcefully pointed out in his comparison of the citizen-centred design in different applications: What is crucial to understand is that digital applications are therefore not a solution to the pandemic in and of themselves, but rather need to be connected with a health strategy of testing and a citizen-oriented design. If not, the applications may create a whole series of additional problems in addition to the lack of free movement by depriving citizens of other liberties, as well as by contributing to the progressive installation (by ratchet effect) of a control society that jeopardises privacy and fundamental rights by default.
That is why it must be stressed that current proposals for tracing applications must in no way be judged on their technical performances (including privacy design only). They must be judged by following a “citizen design” protocol, which in this case must start from a genuine clinical definition of health objectives. If we are not careful, Covid-19 applications could become routine surveillance measures for the purpose of overseeing compliance with the rules, including progressive deconfinement of a part of the population, as well as punishment for patients who do not declare themselves. This would be a most successfully failed ultra-solution in the sense that Watzlawick has proposed. In this case, these ultra-solutions would add mass surveillance of populations and discrimination to the suppression of individual freedom of movement.
Conclusions
Those countries which were least prepared for the pandemic have sought the most refuge in digital techno-solutionism, trying to divert attention from the catastrophe they failed to prevent. Tracing applications have also become quite popular in countries which hope to step up their digital sovereignty by propping up their own technological industry. Strategic communications companies have instrumentalised these sentiments in order to help governments convince the public to participate in the collection of their health data in the name of national health. To slow the pace of, or to prevent, a control society, it would be necessary to at least create strong sunset clauses and oversight mechanisms for the entire process, with the possibility to stop the collection of data after the emergency.
We must order the destruction of sensitive data that are stored during the pandemic and refuse that the process is repurposed for other reasons. The laws and decrees which are legitimating these applications certainly have to rely on existing oversight mechanisms: data protection, ethics guidelines, sanitary codes, and/or ombudspersons. This is badly needed to avoid having only ad hoc technical oversight. The various governments genuinely ready to pursue a citizen-centred design and to safeguard privacy must also enhance the collaboration between these oversight structures and reinforce their powers, including, for example, sanctions against data brokers should they be using these data for commercial interests.
Therefore, what matters in the long-term is which digital platforms will manage the applications after the pandemic has been brought under medical control, which public-private assemblages will have consolidated their power and profits, and which uses of health data will continue after the crisis. There is every reason to believe that the implementation of applications whose data are certainly anonymised during the process, but are later sent to interoperable systems, will finally connect the health data of individuals with police and border-crossing data identification systems. This is likely to integrate different spaces, from the most local (municipalities) to the transnational spaces of the global North (airports, train stations, etc.). This dystopian scenario needs to be rejected without ambiguity by organising a reinforcement of human rights legislation both internally and at the EU level.
De-anonymisation mechanisms through cross-referencing of data are increasingly effective, and data brokerage companies have shown that they can already know our private information and personal thoughts without necessarily needing our name (Cambridge Analytica). If sensitive health data is circulating and is a source of profit, then it is likely that the routine use of such applications for surveillance purposes, and the dependence of free movement on authorisation, is a serious risk for democratic societies as a whole. In Europe, it will be difficult to find national solutions to limit the power of the transnational guild of digital technology professionals. Our greatest hope may come from listening seriously to the warnings of constitutional and human rights lawyers, as well as pushing the European courts via strong civic mobilisation to have their say in this debate. After all, it concerns the very core of our democratic societies.