I have been reading up on the privacy issues around the NHS contract tracing app for a bit now and this is the clearest and most useful thing I have read from the people developing the app. Lots of detail in there, will tweet some in this thread /1 ncsc.gov.uk/blog-post/secu…
First thing you need to understand is the difference between a “centralised” and “decentralised” app design /2 
The NHS have decided, differently to a number of other countries’ health systems, that they will go for a partly centralised approach - The final bullet point is why, They see the public health insights a centralised app will provide as worth the added risk to privacy /3
I think I have got this right, the focus on “risky contact”, I.e. people who have symptoms but haven’t yet been tested, requires a centralised approach because you want to be able to tell others who have been alerted that there has been a negative test and to stand down
Here’s the really important bit - how vulnerable is the centralised app to attacks on individual privacy. The article is pretty clear on this but unless you really understand the tech it’s difficult to critique these statements /5
The article is worth reading. Doesn’t shy away from fact that a centralised approach generates privacy risks. But also rightly accepts there is a balance to be struck between effectiveness and privacy. The key question is whether people will trust a centralised app /6
Ultimately you have to remember this is really new technology and no country has more than a few weeks of experience of using it, and every country seems to be doing things slightly differently. So it's kind of the Wild West, but potentially worth it if it helps ease lockdown /7
The app will succeed or fail based on the number of people who choose to use it. Has to be upwards of 60%, as I understand it, to suppress the virus, but Matthew Gould of NHSX told the JCHR that lower adoption could still be helpful in tracking the virus /8
And the slight paradox is that the NHS and govt have to convince people the app is safe and secure to increase take up, so making it more effective (because the NHS has more data centrally) could make it less effective (because less people trust it) /9
... and that is the delicately balance the NHS have to reach for this to work. Ultimately, the more open and transparent the NHS are with how it works and any potential vulnerabilities the more likely the public will trust them. So I comment Ian Levy's article in that regard /10
As to human rights, it's pretty much the same exercise as it comes down to right to privacy which can be interfered with if that interference is a lawful, necessary and proportionate way of protecting public health. Proportionality requires similar balance of risk v reward /11
If you can stand more technical detail, lots of interest in the technical paper which is linked in that article. Again, clearly written and useful even for non-tech people ncsc.gov.uk/files/NHS-app-…
And here is an interesting rebuttal piece from @TheRegister's @kierenmccarthy. Basically says there are major effectiveness (bluetooth not working well in background on either iOS or Android) and privacy (central database not being secure) issues theregister.co.uk/2020/05/05/uk_…
... and also trust. That is something which keeps coming back to me in comments, why should we trust the government (especially this one with Data Farmer Dom) with our data. I think the answer to that is guarantees in primary legislation, tough independent scrutiny, transparency
... of course other option is no app, or no app unless it uses Apple/Google decentralised model. 'no app' is I think politically unrealistic, but latter option of going back to Apple/Google decentralised option & losing the chance for the NHS to use the data could gain traction
... there is a pretty hilarious irony here (if we are allowed hilarity in these times) that the 'Apple Google' option is the privacy protecting hero on the white steed! How times change. I suppose at least it ain't Facebook...
A further update on effectiveness
Matthew Gould (NHSX), Information Commissioner and others' evidence before the Joint Committee on Human Rights on contact tracing - transcript now up. Some interesting bits from Gould's evidence below committees.parliament.uk/download/file/…
