Dr Iulia Cioroianu is a Prize Fellow in the Institute for Policy Research (IPR) at the University of Bath. Anastasia Dal is a final year student in the Department of Social and Policy Sciences at the University of Bath.
In this blog post we map the current debate around the development of COVID-19 digital tools and applications. Using computational methods to collect and analyse more than 18,000 articles published online on this topic in the past six months, we identify the need to protect individual privacy as one of the main areas of interest. The debate points to a demand for interdisciplinarity and tripartite collaboration between governments, the private sector and the public towards designing and implementing a regulatory framework that would ensure the protection of individual data.
We argue that Citizen Science and Responsible Research and Innovation, which are currently mostly absent from this debate, can provide the conceptual and methodological framework which can be applied not only during the current pandemic, but also going forward, such that concerns related to privacy, security, and the ethics of data and algorithm use are addressed without suppressing the impetus for employing large-scale technological solutions for social good.
Overview of COVID-19 related digital tools
The pandemic has generated a need for increasingly complex digital tools which allow us to monitor the evolution of the virus, identify and protect vulnerable categories, and shift many of our activities and interactions online. More than 50 mobile apps have been developed, trialled, and adopted globally, most of them focusing on contact tracing and health monitoring, and in recent weeks also on adapting existing technologies to mitigate the social and economic effects of the pandemic.
Contact tracing apps have been discussed as an alternative to manual contact tracing - the most common response measure to contagious diseases. But the articles and opinion pieces included in our analysis point to the fact that most contact tracing technologies have not properly addressed multiple issues surrounding individual privacy. According to a recent study, only 16 out of 50 existing contact tracing apps indicate that they encrypt and anonymise users’ data. And these security measures have been seen as not sufficient - recent machine learning research shows the ease with which individuals may be re-identified from anonymised large-scale datasets. Thus, it matters how much data about individuals is collected, and it is important to ensure that only the necessary information is processed.
However, just how much data is necessary for digital tools to serve their purpose is a difficult question. The less data is shared, the more difficult it is to accurately determine whether users have been exposed to the virus and the less information is available for research. Following a short period of initial excitement around the use of technology to manage the pandemic, this trade-off quickly became a highly politicised and polarised issue. Coupled with a range of technical difficulties, it resulted in disparities in public interest and willingness to use the newly developed tools across different countries.
A proposed solution to the challenge of ensuring a sufficient number of downloads for COVID-19 tracing apps in order for them to be effective has been to embed the contact tracing function into apps that most people already have. Apple and Google have partnered to deliver Bluetooth-based technology, which has also been discussed as a more secure and privacy-protective option than GPS-based contact-tracing technologies. However, this technology is also more susceptible to inaccuracies. The case of the NHS digital tracing app and its failure to take-off has been widely debated as an example of the decisions and compromises governments have had to make, pointing to broader issues within the field of large scale app development for public good.
Wearable technologies have also been discussed as a solution to mitigating the economic impact of the pandemic and restarting the economy. For instance, an Israeli company is repurposing their ankle monitoring devices for tracking incarcerated or criminally convicted people. Experts have warned that this measure is unnecessary and may be used to normalise the use of coercive police technology. Wearable technology has also been developed for businesses who wish to track their employees, and real-time location tracking devices are already commercially available.
A common theme across the surveyed articles are concerns about governments’ use of the pandemic to exacerbate mass surveillance, divert more funds towards the development of surveillance technology, and to extend these measures beyond the pandemic. Some countries, e.g. Russia and China, have been using facial recognition technology to track “high risk” citizens and people who do not comply with social distancing measures. Rapidly developing amid the pandemic, the technology can now measure people’s body temperature and identify people wearing face masks.
Some technologies have explicitly identified individuals diagnosed with COVID-19, which has led to concerns about social stigmatisation and discrimination as well as economic stigmatisation of small businesses. Facial recognition technology has long been a source of public concern due to its potential bias and inaccuracy, and its propensity to discriminates against marginalised groups. The media and public backlash has led to a drawback from both governments and private companies - Microsoft now declines to sell their facial recognition technology to police, while IBM “will no longer offer general purpose facial recognition or analysis software”.
Mapping the COVID-19 digital tools discussion space
The use of digital tools to handle the pandemic has been intensely debated in the media, and articles about COVID-19 digital technologies drew high levels of social media attention. Hundreds of articles and opinion pieces on the topic have been published daily for the past months, and they have been shared millions of times on Facebook. Using Media Cloud, we have identified more than 18,000 online articles mentioning COVID-19 and digital tools. We then processed them to extract and analyse the text, as well as their social media reach.
Figure 1 shows the English-speaking online news sources which were most active in this debate and generated the highest levels of social media engagement in the first half of 2020. Articles published by the New York Times on the topic have been shared more than a million times on Facebook. A mix of American and British online news sources and press agencies dominated the debate, along with a couple of sources covering Asia and the Middle East. A surprising presence in this space is Visual Capitalist, an online news publisher producing high-impact data-driven visuals, showing the importance of communicating information in a format that is easily accessible to the general public.
The debate intensified between April and May, coinciding with announcements about the use of technologies by multiple governments across the world, and peaked at the beginning of July, when some of the major issues with implementing large-scale digital tracing emerged. Privacy emerges at the most salient theme in the articles about the use of technology to tackle the COVID-19 pandemic published in the past six months. While technical and factual pieces dominated the debate, a large share of the articles published also mentioned concerns about individual privacy.
However, Figure 3 shows that articles that discuss privacy issues receive far less social media attention than articles that do not address these issues, suggesting that even though in opinion polls the majority of the population expresses concerns about the use of digital tools and tracing apps, the public is not actively involved in conversations around this topic, and does not actively engage in debates about the implications of this technology.
Differences also exist in the extent to which major news sources cover the issue of privacy, and the engagement that their coverage generates on social media. Relative to their overall social media reach, the Associated Press, LA Times and Visual Capitalist have been the most active sources in this debate, and have generated comparatively large levels of social media engagement. Surprisingly, despite the fact that the New York Times published a large number of articles on the topic of COVID-19 technology, and is also the source with the overall highest levels of social media engagement, their articles covering privacy issues received lower overall levels of social media engagement compared to those of the Associated Press.
A characteristic of this debate is the fact that technological solutions are presented as a trade-off between the privacy of individuals and their health, balancing public health protection with safeguarding civil rights. Many of the surveyed articles emphasise that the emergency of the pandemic should not enable political actors who have an interest in creating a surveillance state to do so, and point to the fact that the legal and ethical framework for collecting, storing and processing private information is still missing.
The pandemic has thrown the long-standing tensions around surveillance and privacy in digital tool development under the spotlight, presenting an opportunity to establish the much needed legal and policy framework for democratic and transparent design, implementation, and oversight of large-scale technological solutions. From a government perspective, this can also become the driving force for innovation and collaborative action aimed at developing robust democratic systems around large-scale technological solutions for social good.
The need for public deliberation and co-creation
Yet an aspect that is blatantly missing from the articles surveyed is the need to increase public deliberation around the tools that are being developed and the ways in which they are being implemented. To address the threats to democratic freedoms posed by digital tracking technologies, and the resulting loss of trust, the public needs to be meaningfully engaged in their creation process. It is not enough to survey citizens. To ensure meaningful and inclusive participation, a more deliberative approach must be taken. Meaningful public engagement generates procedural fairness and trust, improving the efficacy of what the state is trying to achieve.
Framing the debate only around the dichotomy between privacy and surveillance in this context ignores lengthy academic conversations that have been carried for years around these topics, mainly in the subfields of Citizen Science and Responsible Research and Innovation (RRI). These two conceptual and methodological frameworks have evolved in recent year in response to the increasing need to engage the public in the creation of knowledge in science and public policy in order to preserve public trust in the research process, its output and the technology used and resulting from it.
RRI and Citizen Science principles are very closely aligned with the calls for more careful ethical consideration of COVID-19 digital tools, yet none of them are expressly mentioned in the thousands of articles on this topic that we surveyed. The two approaches are based on a set of similar principles, such as the need to ensure that scientific development involves all relevant stakeholders throughout the research and innovation process. A standard design would involve citizens working with scientists, policymakers and innovators to set the agenda and agree on the way scientific discoveries are implemented, in light of their broader implications. Yet none of these steps have been taken when developing existing COVID-19 digital tools.
The accelerated pace at which existing tools have been created has forced governments and private actors to revert to the simplified principles of citizen participation, instead of deliberation, and crowdsourcing, instead of co-creation.
Yet carefully designed, properly implemented citizen science is particularly useful in contexts where a fast evidence-based response is required, such as during a pandemic, because it mobilises large groups to produce evidence on which to base policy decisions. In the absence of an alternative set of incentives to participate, citizen engagement in pandemic response activities depends on their collective trust in the necessity of compliance and on the emergent social norms that alter individuals’ behaviour for the good of the collective. This is particularly important in countries where trust in the state and the way it had been handling the crisis has been low or has dramatically decreased over time, such as the UK. Developing digital tools based on citizen science principles offers an avenue for rebuilding public trust through citizen empowerment and meaningful involvement in the decision-making process.
All articles posted on this blog give the views of the author(s), and not the position of the IPR, nor of the University of Bath.