The COVID-19 outbreak is proving an interesting time to be a data protection practitioner. There seems to be a new article each day about the next exciting app which promises to use data to help manage the crisis.
This post focuses on two particular propositions that pose interesting data protection considerations. It also flags the wider issues that developers should bear in mind when trying to respond to this unprecedented crisis.
It was reported on 31 March 2020 that the UK government is actively set to develop some form of contact tracing app in the near future. This follows successful app-based contact tracing in Singapore and South Korea. Led by NHSX, the innovation arm of the NHS, the app will leverage Bluetooth to identify individuals who have been in close proximity to each other, storing a record of that contact, and providing a mechanism through which an individual can be notified if they have been in close proximity to someone that tested positive for COVID-19. Given the anticipated use of Bluetooth, it is possible that NHSX may leverage Singapore’s TraceTogether app which used the same technology, the code for which was open-sourced by the Singapore government last week. TraceTogether was widely praised for collecting the bare minimum of data despite the extraordinary circumstances at hand.
The success of any tracing app will depend on a critical mass of users downloading it. Concerns are already being raised about whether private entities might require either employees or customers to use the app, to show they have not been in contact with infected individuals. It will also depend on a comprehensive testing regime to ensure that those who are symptomatic are tested quickly so that the notification can be sent appropriately quickly. Similarly, swift testing may help avoid people being unduly required to quarantine themselves having been in contact with someone with minor symptoms which do not turn out to be COVID-19.
It is interesting to note that initial statements from NHSX suggest that contacts will be stored on users’ phones, with notifications sent via the app after a suitable delay to avoid identification of the infected individual. It is not currently intended that the data would be sent regularly to a central authority, which may give comfort to people concerned about their privacy. Additionally, NHSX has indicated that it intends to appoint an ethics board to oversee this project.
COVID Symptom Tracker
ZOE, a health and data science company, in conjunction with Tim Spector, a genetic epidemiology professor at Kings College London, have created an app called ‘COVID Symptom Tracker’ that allows users to self-report potential symptoms of COVID-19, even if feeling well. The aim is to use this data to track the progression of the virus in the UK, and potentially identify high risk areas.
The app is based on consent, both to the data processing and to potential transfers of personal data to the US. Data is collected for the following purposes related to COVID-19 including: (i) better understanding the symptoms; (ii) tracking the spread of the virus; (iii) advancing scientific research into links between patient health and their response to infection with the virus; and (iv) potentially to help the NHS support sick individuals. Whilst at an initial glance this seems like a reasonably narrow set of processing purposes, you could envisage a surprisingly broad range of activities which might fall within these categories, including specifically tracking individuals.
Data protection considerations
When it comes to processing personal data, the post-GDPR mantra is increasingly ‘Just because you can, doesn’t mean you should’. The principles of fairness, transparency, purpose limitation and data minimisation in particular will require serious consideration to ensure that the proposed data usage is justifiable.
Whilst the Secretary of State for Health & Social Care Matt Hancock recently tweeted that “the GDPR does not inhibit use of data for coronavirus response”, this may not necessarily be aligned with the ICO position that the GDPR is still in full force, despite the fact that the ICO may take a pragmatic approach where necessary. There are certainly lawful routes to using personal data to fight COVID-19, but this should be done based on clear reasoning and analysis.
With that in mind, the following key considerations may assist when evaluating whether or not to use personal data in the context of COVID-19:
- be confident that you have an appropriate lawful basis for processing the personal data. Remember that both vital interests and substantial public interest are very high bars to satisfy. Likewise, legitimate interests always needs to be balanced against any potential impact on individuals’ rights and freedoms;
- do not use personal data for extraneous purposes. You should aim to keep your processing purposes as narrow as possible for the stated aims, and be conscious that any attempt to use the dataset for non COVID-19 related reasons might be seen as acting in bad faith. Similarly, the collected data should be limited to what is strictly necessary for the processing purposes. Avoid the temptation to collect additional categories of personal data because they ‘may’ be useful in future;
- the potential volume of data processing, and categories of personal data being anticipated, suggest that in relation to many of the COVID-19 related apps a data privacy impact assessment should be undertaken. These should be completed carefully and not rushed for the sake of getting an app into the live environment;
- consider who personal data is shared with, and whether sharing a full dataset is strictly necessary. It may be possible to anonymise personal data such that the recipient only receives fully anonymised data, which may help manage data subject concerns about where their personal data might go. Remember however that true anonymisation is difficult and the pseudonymisation alone does not take data outside of the scope of the GDPR;
- given the potentially higher risk processing that is taking place, it is important that data subjects understand how their personal data will be used, and who it may be shared with, particularly where they are giving up unusual freedoms such as in the context of tracking. Data controllers should aim to go above and beyond to ensure their fair processing information is clear and easy to understand, so that individuals have good expectations of how their data will be used;
- if and when relying on data subject consent for any processing, it is likewise important to ensure that the individuals understand exactly what they are consenting to. Now more than ever it is vital that consent is specific, freely given, informed and explicit when dealing with sensitive health data;
- personal data collected in the context of COVID-19 is generally required for the specific aim of managing the outbreak of the virus or its effects. This may mean that it is not necessary or appropriate to retain this personal data once the virus has been controlled and life returns to normal, depending on what has been communicated to data subjects; and
- holding larger volumes of personal data, or special category data, potentially represents a higher security risk and there may be increased cyber attacks on the dataset. Ensure that you have appropriate additional security measures in place where necessary.