The CDEI recently hosted a virtual roundtable with people from nearly 10 different countries who are working on contact tracing apps. The discussion focused on public trust and how it can be built while working at speed to develop and deploy a new technology. We wanted to explore the different approaches being taken and understand common themes.
Since March the CDEI has been working with NHSX as it develops a COVID-19 contact tracing app. In particular, the Centre supports the new Ethics Advisory Board established to provide constructive challenge to the app development team.
We are particularly interested in contact tracing apps as they provide an example of the potential for data driven technologies to be used for public good. Yet unless they are developed and deployed ethically, public confidence may erode and the opportunity might be lost - not only when it comes to this specific technology but in future government attempts to use data and AI as well.
Our international discussion highlighted that the different countries represented are all facing similar challenges. The speed of development means that working transparently and enabling scrutiny is not straightforward. New decisions are taken every day, and sometimes changed a day later as new evidence or technical challenges emerge. Explaining this to the public without undermining confidence is hard - particularly at a time when people want to be reassured that their governments have the crisis in hand.
Others were mindful of the need to address concerns in relation to the privacy and security features of the app, without making guarantees that might need to be reversed in the future, or that might reduce the effectiveness of the app and undermine its value. A particular challenge comes in collecting additional data through the app and also enabling it to be used for research. While this might make the app a more valuable tool by extending its use beyond contact tracing, some countries have ruled this out to address concerns about privacy and potential function creep.
Many countries had undertaken user research and polling and found that the majority of people said that they would consider downloading and using the app. It was noted that this did not equate to people definitely doing so, and a technology that depends on mass take-up would require a clear communications strategy. It is also likely that the views of the public may change as perceptions of the risk of COVID-19 evolve. Indeed, providing people with the confidence to download it is only the first challenge, for it to be effective it must also be used.
Participants also spoke of the importance of context. In most countries, citizens are living under lockdown with their civil liberties heavily curtailed. Perhaps people are prepared to trade-off an element of privacy if it enables them to begin to return to normal and also help to keep their communities safe. Such trade-offs need to be clearly articulated while also ensuring citizens benefit overall. An app encroaching on people’s privacy while providing little value in tackling the spread of COVID-19 would be ethically dubious.
People’s confidence in the app and likelihood to download it largely depends on whether it is viewed as effective. A number of countries highlighted research into public attitudes which found that, for many people, effectiveness was felt to be of greater concern than privacy. But working amid so much uncertainty means it is hard to know what will work. Scientists are confronting a new virus that is not well understood. This is a new technology that has not previously been deployed at scale. And populations are likely to be nervous about returning to some form of normality. There are few easy decisions. Elements of (evidence based) experimentation are inevitable. Participants in the discussion recognised that the app is not a silver bullet and must form part of a package of measures (including manual contact tracing) designed to stop the spread of the virus.
It was notable that countries appear to be coming to different decisions as to how to design their contact tracing apps. For example, some countries collect phone numbers from app users so that they can be contacted by human contact tracers. Others have taken the view that in order to build trust and increase likely usage it is better not to collect such personal information. Learning from experiences in other countries will help to inform decisions here.
Judgements must be made and ultimately political decisions need to be taken. However, such decision making can be guided by an ethical approach, identifying the trade-offs and endeavouring to reflect the reasonable expectations of citizens. The CDEI has worked with the new Ethics Advisory Board at NHSX to identify a set of core principles to guide the development of the app. Over the coming weeks we will publish further blogs to explore these in more detail.
|Value||There must be good reason to believe that that the app will provide sufficient net-value back to the citizen or society as a whole so as to justify its introduction and any adverse consequences for individuals.
There must be a compelling value proposition to users.
|Security (and privacy)||Data sharing and storage should be secure. The data collected should be minimised and protected as much as possible, so users’ privacy is preserved.|
|Accountability||There must be a reliable and understandable decision-making process in place to manage the app - with clear accountability, particularly with regards to introducing new functionality, data collection or use cases.|
|Transparency||A strong and demonstrable commitment to transparency enables scrutiny and helps to build trust.|
|Control||Users should be able to see what kinds of data are held about them so that they can understand how it is impacting on decisions.|
About the CDEI
The CDEI was set up by government in 2018 to advise on the governance of AI and data-driven technology. We are led by an independent Board of experts from across industry, civil society, academia and government. Publications from the CDEI do not represent government policy or advice.