There are reports that the COVID-19 contact tracing app will be launched in England and Wales on 24th September.
Professor Pete Fussey, Department of Sociology, University of Essex & Research Director of the ESRC Human Rights, Big Data and Technology Project, said:
“There are three key areas about the app which need to be considered: effectiveness, purpose limitation (explained below) and governance, oversight & responsible use.
“The effectiveness of the app depends on how the different elements of the system integrate. The original plan to have a centralised app (presumably so govt./NHSX/vendors could retain data), which, as the experience of other countries has shown was always doomed to failure. The google-apple approach is ‘decentralised’, i.e. the data remains on people’s handsets. This is more straightforward to implement and has greater privacy protections.
“There appear to be several elements at play: First, to work, the app would have to be companioned by an effective, reliable and quick testing infrastructure. This is yet to be established. Also, while people might have to scan a QR code, another stated element seems to be that the technology relies on Bluetooth ‘handshakes’ with other devices in the vicinity. This has been considered a particularly weak part of the chain: Bluetooth can connect through walls and other partitions. This means people in high density accommodation (who also are most likely to be employed in precarious labour) have increased likelihood of being be misidentified. This instances are also likely to occur in high density urban areas where health services are most likely to be under strain.
“Relatedly, there are significant issues of public trust, with an overlap between those most vulnerable to the virus (e.g. low income, BAME etc.) least trusting of state agencies, which could affect levels of opt in and protection. Low level of adoption has been an issue globally.
“All this is to say, that there needs to be a threshold of effectiveness for it to be worth using.
“Purpose limitation means that the collection and processing of personal data must have a clearly defined application and can’t be used for something outside of this purpose. It is not clear what the limitations will be placed on the information collected by the app. The data protection impact assessment for the Isle of Wight trials demonstrated significant shortcomings and ambiguity in this regard. In addition, it is not clear how what functionality is to be built into the app down the line, e.g. in enforcing self-isolation etc. In other countries there has also been the potential to reverse identify ‘anonymised’ data.
“There is very little effective regulatory oversight of the app. The Information Commissioner’s Office has some notional oversight of some aspects of the management of data but have so far only listed general expectations. While ‘ethical oversight’ is an insufficient mechanism to regulate a surveillance measure as potent as this, the apps ethics panel was disbanded very soon after its formation. One analysis of the human rights implications of contact tracing apps has been articulated by the Parliamentary Joint Committee on Human Rights, which also includes reference to our ESRC project team’s detailed consideration of this technology: https://committees.parliament.uk/writtenevidence/3973/html/
“Overall, an effective contact tracing app could be an excellent tool for addressing the pandemic. Yet it is important to resist the seduction of what the technology promises without thinking through the consequences of its implementation. Rushing out a nascent app without sufficient technical and oversight infrastructure surrounding it is unlikely to be effective or mitigate the range of potential harms. This is why I feel calls for a public duty to download the app or framing the issue as privacy vs public health are problematic.”
Dr Scott McLachlan, Postdoctoral Research Assistant, Queen Mary University of London, said:
“There needs to be improved transparency and clarity about the number of users, rather than just the number of downloads. We know from other attempts (including the Isle of White rollout of the first NHS App) that you can have 300,000 downloads but find that only 30,000 people can actually install or use your app… and that only a percentage of those 30,000 people will go on to use the app for longer than a day or two. It is this last number that are the real qualified users. For many of the global contact tracing apps that have been implemented, the number of qualified users has been observed to be fewer than 25% of those who downloaded the app. For example: this could mean that of the 1.3million who are said to have downloaded the Irish Celtic COVID app less than 300,000 may be qualified users of the app after as little as a few days. We need more information in order to have a better estimate of the number of users.
“We must ask why are we still harping on about developing and deploying these hugely expensive and proven ineffective contact tracing apps for COVID when the countries the politicians keep referring to (like Singapore) have long since realised their apps don’t work? We know that voluntary adoption of COVID contact tracing apps has been extremely low. Australia and Singapore have had less than 25% uptake and the Australian app only led to ONE confirmed case of the virus from using the app. Norway abandoned their app around the same time the UK abandoned the NHSx app, because they found that while 1.5mil people downloaded the app, only around half had installed or were using the app, and that this equated to only 20% of the adult population. The Australian Govt have abandoned the idea that a contact tracing app will do anything effective for the general population, acknowledging that a single app is never going to work properly on all versions of all phones that are ‘out there’ in the community.
“The recent Apple update (13.7) that puts the contact tracing API directly into the iPhone’s OS only works on phones newer than the iPhone 7 (so the 8, X and 11) and leaves out everyone with an iphone before 2017. The same is true for the Android ecosystem, in that anyone with a 2016/17 android phone running Nougat (Android 7) or below remains unsupported by Google’s Exposure Notifications API. Together, this means there is no support for around 50% of all smartphones currently in use. How can you get 60% of the population to use a smartphone app when only 79% of our population have smartphones, and around half of them are unsupported? That means at best you can only hope for somewhere around 40% of the overall population to ever be running the app – an unlikely outcome as it would require 100% of those who can install the app to be running it. An impossible goal and certainly one much lower than the 60% population-wide usage the govt has maintained that they need.”
“Many third parties, such as coffee shops and other businesses, have been running COVID tracing apps that use QR and other codes to record patronage contacts for some time, and as such the Government’s App may be seen as late to the party and a little ‘me too’. It also undercuts the political claim that what they are supplying is cutting-edge technology.
“The QR code route is simply another way of collecting citizen’s location data while claiming privacy and security by not exposing or collecting our GPS data. Given that the QR codes will be tied to a particular location, the Government will know where and when you were, and for what it is worth, may as well be tracking you with a GPS locator chip.
“We are still not told how they will deal with the multitude of false positive ‘contacts’ that all contact tracing apps have been creating, nor how much tracking all those false positives is going to cost the taxpayer.
“Some groups of the public are already starting to show COVID fatigue. They are reacting negatively to yet more news reports about COVID and say they just want to get on with life. I believe this fatigue means individuals are less likely to want to download the app, or to persist in using it for the prolonged period of time necessary to achieve any meaningful outcome.”
Prof Rowland Kao, the Sir Timothy O’Shea Professor of Veterinary Epidemiology and Data Science, University of Edinburgh, said:
“This contact tracing app is a useful addition to traditional contact tracing methods, in particular because it allows for the rapid identification of casual contacts for whom a potentially infected person may not have contact details. The app being adopted here is the same as the one that has been used in Ireland since July and so should be operational without any major issues expected. It does not require central administration of personal details, though it may be possible to reverse engineer individual information for someone with the contact information, should sufficient other details also be available.
“The app is not a substitute for traditional contact tracing – first the app requires both ends of the contact trace to have the app downloaded (meaning that even high adoption rates may still miss many contacts) and second, there will be many instances where personal contact and ‘soft skills’ will be necessary to ensure that quarantine and testing requirements are well understood and followed. Most importantly, even where tracing rates are high, COVID testing must not lag behind, either in terms of local availability of tests, or speed at which test results are received.”
All our previous output on this subject can be seen at this weblink: