The UK digital contact tracking app is almost here and even the organization behind it admits it’s still a work in progress that has the potential for ‘unintended consequences’. It can also be a disturbing example of how digital contact tracking will turn out during the coronavirus pandemic in the United States as an uneven network of state-wide contact tracking programs begin to roll out.
For a contacts tracking app to really work, it needs to be widely applied. So if an app doesn’t have the trust of its intended users – due to privacy issues, functionality issues or even because of it ask to be over legality – people will not use it. On the other hand, an app that preserves privacy too well does not provide public health authorities with enough information about the spread of the virus for a good response. The UK joins the US, France, Germany, Singapore and many other countries to find the difficult balance of creating an app that is both useful and useful.
Trials of the UK National Health Service (NHS) app have started on the Isle of Wight, an island off the south coast of England with around 140,000 residents (who appear to have mixed feelings on detecting guinea pigs in the UK). Government employees were given access to the app on Tuesday and the rest of the island population can use it on Thursday. The results of the pilot program will determine if and when the app is rolled out across the country.
The app is somewhat controversial because of the NHS’s decision not to use the Apple-Google exposure notification tool, which released in mid-May. Like many others, both the NHS contacts tracking app and the Apple-Google tool rely on exchanging Bluetooth signals with nearby phones. But while the Apple-Google tool severely limits the amount of data exchanged, the NHS wanted more information and control than Apple-Google would allow, leaving it alone.
“An app that offers fantastic demonstrable privacy but does not help stop the spread of the disease is not a useful tool,” Ian Levy, technical director of the UK’s National Cyber Security Center, who helped develop the app , said in a blog post.
First, the NHS wanted a centralized system, storing the Bluetooth keys of both infected people and those they came into contact with in a database maintained by the NHS. The Apple Google tool is decentralized, so only an infected person’s anonymized Bluetooth keys are uploaded to a server. Details of any agreements with devices that have come close to the infected person remain on those devices’ phones without being uploaded to a central server.
This is more private, but it also means that public health authorities receive very little data. Some, like the NHS, say they need more data to study how the virus spreads, but some of this data, such as zip code information, can be used to identify specific users, which can cause privacy issues. Furthermore, the NHS is not excluded add more identifying information to the app in the future, such as collecting location data.
The NHS not to have the best reputation when it comes to keeping patient data safe and private, making it more difficult for citizens to trust that it can protect the data it gets from the app. Matthew Gould, head of the NHSX, the NHS technical and data arm, recently admitted that he could not give Parliament “a definitive list of exactly who would have access to the data”, which is not particularly reassuring.
There are issues with the NHS app other than privacy issues. Without the Apple Google tool, Apple devices cannot exchange keys if the app is not open, which could reduce a device’s battery life. The Australian government ran into this problem with the recent release of its own centralized tracing app. The NHS app is also much more likely to generate false positives because users can report symptoms themselves, triggering alerts to their contacts even without confirmation that someone has the corona virus.
The UK isn’t the only country that believes the Apple-Google tool has a little too much privacy and too little control over health authorities. France similarly decided that a centralized system is more important than the technological benefits of the Apple-Google tool (the French app will start testing next week). But other European countries, including Germany, Italy and Switzerland, support the decentralized approach, although it is not yet known whether any of them are using the Apple-Google tool.
In the U.S., no individual states or the federal government has ever committed to using the Apple Google tool. Some states, such as Utah and North Dakota, already use their own digital tracking systems and see limited success, given the reported North Dakota app 3 percent acceptance rate. Several others, including Colorado and Louisiana, told Recode they have no plans to use the tool in apps they develop.
What remains to be seen – and what many experts have expressed There is considerable doubt as to whether any of these digital contacts tracking apps, whether centralized or not, will be widely used upon release. Singapore’s experience, where only one fifth of the population downloaded the TraceTogether app does not bode well. a often quoted report suggests a minimum adoption rate of 60 percent of the population for a contact discovery app to be effective.
When participation is voluntary and opt-in, the responsibility lies with the health authorities to make their apps as tasty as possible for the general public. For many, that includes those apps that provide minimal data on their contacts and health status. A centralized database doesn’t do that, and that may be too much of an obstacle for this NHS app to do what the agency hopes.
Open Sourced is powered by Omidyar Network. All Open Sourced content is editorially independent and produced by our journalists.