Trialled this week on the Isle of Wight, and soon to be rolled out to us all: an app for our smartphones will record who we come into contact with, so that if either we or they are discovered to have covid-19, the other party can be informed and take action.
Don’t confuse this with the Covid Symptom Tracker app, which you should have already. This allows you to anonymously record your symptoms daily to assist in research on the spread of the disease and on how symptoms begin to manifest.
The new app will use the bluetooth radio feature on smartphones to track proximity to other phones (and therefore people). This is a feature that phone OS companies (Google and Apple) have restricted because there has not been, hitherto, a legitimate reason for an app to track this kind of information. Use of bluetooth by apps in this way is largely restricted by design to apps in the foreground, i.e., showing on the phone screen, while the screen is on.
Google and Apple have indicated that they will allow bluetooth to be used in the background for this provided that the process of matching up the (anonymised) ID of a known case to your (anonymised) ID happens on your phone and not on a central server. This is the ‘decentralised model’. The UK government has decided that this isn’t good enough and it wants to know the pattern of risk of transmission, for research and monitoring purposes, and sees benefits from central control of notifications (more here). It is therefore going for a centralised model. It claims to have found some way round the limits put in place by Google and Apple.
There’s more detail on issues around the two models at the Register, here.
The government is making a mistake. While the IDs are anyonymised, there is still potential for somebody with access to the servers and to other data, to crack this anonymity and operate mass surveillance of contacts between individuals. This is the privacy concern, and this will understandably put many people off using the app. That is not what we need. And hacking your way round barriers in an OS is no way to build a stable, reliable app. There will be downsides to this, whether it is battery, missed contacts, other unreliablity, etc. Most of the rest of the world is following the decentralised model.
Having said all this, I will install the app. Right now my freedom is far more restricted by my not being allowed to travel to visit my friends, etc, than by the theoretical risk that the government might know I was visiting my friends. I might uninstall it sooner than I would a decentralised model app, during the long tail of this disease, but right now we need to know when we, and those we contact, have been exposed to risk of infection. I would prefer a better app, but the best must not be the enemy of the good.
Can we let the government get away with making the wrong call here? Yes. I have the right to protest by not installing the app, but I won’t exercise it on this occasion. Lives are depending on the information that an app like this can generate, not just mine and not just yours, and I can see no justification for not installing it.
* Joe Otten was the candidate for Sheffield Heeley in June 2017 and Doncaster North in December 2019 and is a councillor in Sheffield.
10 Comments
There’s not much point in me using it – I’m not going anywhere or meeting anyone apart from my husband!
The fundamental flaw in this app is that its benefits accrue more to big organisations than to the individual, so that promoting it as a public-spirited duty is wrong. (Like electricity smart meters.) Any central database is by its nature a honeypot: how much would it cost to de-anonymise the data, versus how much profit can be made from doing that? People exposed to the virus would be juicy advertising targets, not just for small fry (like my friend who makes PPE), but also for rich politicians who trade on fearfulness (like Trump and Johnson), and for secret influencers (like Putin and Cummings). Beware mission creep.
There’s an update to the article at the Register, linked above, suggesting that background use of bluetooth is working in the NHS app.
https://twitter.com/jasonkneen/status/1257807637124710403
One of the concerns was that it wouldn’t – that restriction of this background use – would hamper the centralised model. (The decentralised model used an API on the OS that didn’t allow the data off the phone.)
Obviously this is a good sign that the NHS app will work, though if it is because they have found a loophole, it may get closed in an OS update. If Apple/Google dare.
Hi Anthony, obviously the major benefit is not to the individual user, but to the community of people that user might otherwise infect and possibly kill if they weren’t aware of being a transmission risk.
Any benefit to corporations would pale into insignificance compared to that. And I do think you need to substantiate your allegations. Cummings is a bad person, therefore this is a bad app? Really? Is that all you have?
Me too, Mary, although the person I am meeting is my wife! However, we do see our sons and their families, in the case of the one in Manchester, who already works from home, as does his wife temporarily, via Face Time and in the case of the one who lives locally, whose wife works part time in our local hospital and who is currently furloughed, from a distance.
What we are really concerned about is whether we will ever be able to resume our babysitting duties again. In happier times my son and his wife have relied on their combined parents for a while now to help getting their small children to school twice a week, picking them up and feeding them before returning them home for bath and bedtime. When they reopen our local schools fully, they and we are going to be faced with a real dilemma unless we get positive guidance. No app is going to help us, I’m afraid. “What about paying for childcare?”; you may ask. Fine if a) you can get it and b) it doesn’t take up all of my daughter in law’s admittedly part time salary. OK, there are people far worse off than our family; but that doesn’t really solve the problem, does it? I sometimes wonder whether we grandparents should be reclassified as ‘key workers’, for the contribution we make to keep the economy going?
One of my concerns is that the app is wide open to malicious use. Let’s say a user stands in a queue for a supermarket or bank for the typical half-hour that it takes to get in. Even though people are standing at around the nominal two metres separation, in practice some will be slightly closer (I have seen just how bad people are at judging distances) and will record as proximate to the user. Suppose now the user, self-declares to be suffering symptoms; the others adjacent to the user will be instructed to self-isolate for 14 days. There doesn’t seem to be any check that the self-declared sufferer has really got the illness.
The UK Government’s version of the Tracker App is flawed. Not only are they trying ti reinvent the wheel, it will not have access to parts of the a phone’s operating system in the way the Google/Apple App does (parts hey don’t give access to for 3rd party App builders) so will not work as well. Worse still, as most of Europe has gone for the Google/Apple App, those rely on the UK Government App may find problems if they want to travel abroad (or even live near the UK border in Ireland).
As with so much of the Government’s response to Covid-19, they have made initial decisions based on ideology and a sense of British exceptionalism only to have to back track later, playing catch up with other countries.
To render the app truly effective, someone instructed to self-isolate needs to be able to secure a prompt test delivered to his or home or conducted there, to avoid unnecessary self isolation. We await to see if the Government will get its act together as far as testing is concerned.
I will not be using it as I do not have a smart-phone. Moreover, as some who has recently moved from the Lib Dems to Labour (in time to vote for Keir Starmer as new leader), I believe that may in the LD party are rabid about personal freedom!
FT reports on speculation that the UK may revert to the decentralised model.
https://www.ft.com/content/d44beb06-5e3e-434f-a3a0-f806ce06576c?fbclid=IwAR3UtlkVmd0DliaCncWMKpKrbpsHAixe70bf_PS6bHq6r7YM38sAkZiVz0I