Lockdown endgame? There won't be one until the West figures out its approach to contact-tracing apps
- Reference: 1587408966
- News link: https://www.theregister.co.uk/2020/04/20/covid19_apptracking/
- Source link:
This isn't inherently novel. It's a tactic that's long been used by epidemiologists, and was featured in the (eerily prophetic) Matt Damon flick [1]Contagion . But unlike previous pandemics, governments can now rely upon the ubiquity of smartphones to trace the movements of individuals. And they are, bolstered by the support of tech companies and homegrown techies.
Apple and Google: a COVID-fighting tag team
As the makers of the biggest smartphone operating systems in the world, Apple and Google, are ideally placed to set the groundwork for any contact-tracing apps by offering developers in the sphere access to more sophisticated APIs. And indeed, that's exactly what's happening.
Earlier this month the two companies, which are otherwise fierce rivals, announced joint plans on this. The goal is to create a standard that would allow government agencies to build apps that track contact and physical proximity with infected COVID-19 sufferers.
This will be done via the Bluetooth radios found in almost every phone. The app will constantly broadcast beacons. When a phone enters range of another using the app, it'll perform a handshake and exchange beacon identifiers.
The app will also record the proximity between the two devices, as well as the duration of contact. If a person is later found to have COVID-19, the apps will then advise anyone who had contact with them to self-isolate or get tested.
Both companies are working to an incredibly tight deadline, with the initial APIs to be released next month. Of course, that won't be the end of the story. Though Apple is able to exert absolute control over its platform, and can push out updates on a whim, it's a very different story for Android.
For Android, updates are made available by developers and carriers. Android phones tend to have shorter update life cycles, too. Apple continues to serve system updates to the iPhone 6s (first released in 2015), but most Android phones fall by the wayside after a year or so. Even the Android One programme only offers three years of patches, which is comparatively paltry.
The end result is that a majority of Android phones will not be able to take advantage of these APIs. Furthermore, it may take a long time to see this patch make its way to individual users as vendors are famed for dragging their feet when it comes to software updates in normal times.
There's another question worth raising: will Android's contact-tracing APIs rely upon Google Mobile Services in any way? If so, that'll be problematic for China's millions of Android users, as well as those in the West who've bought one of Huawei's latest Google Mobile Services-free phones, like the Mate 30 Pro and the Honor 9X Pro. The Register has asked Google about this, and if we hear back we'll update this post.
Academics: We hate to ask, but could governments kindly refrain from building giant data-slurping, contact-tracing coronavirus monsters? [2]READ MORE
Perhaps the most fascinating thing about the Apple-Google partnership is that both companies have a fundamentally different (if not incompatible) ethos. Apple is primarily a hardware company, and makes its money by selling high-margin devices with a user-friendly UI. Google, on the other hand, is a services company, with advertising one of its primary revenue drivers.
With that in mind, how will privacy work within the context of this initiative?
First and foremost, both companies have stressed that the APIs will not hinge upon the collection of location data. Distance between contacts will be recorded by information from the Bluetooth, rather than using GPS.
Furthermore, because contacts will be recorded as anonymous beacon keys, rather than biographical information, there's little concern to individual privacy. If a person is informed that they had contact with someone carrying COVID-19, they won't be able to discern who from the information provided.
Both companies are working on a second iteration that would allow contact tracking to take place within the operating system without an app being installed. This is expected to land in June.
The Singapore solution
One of the few coronavirus "success stories" has been Singapore. Despite its proximity to regional hot spots like South Korea and China, the country managed to flatten its curve early, recording comparatively few deaths to some of its regional neighbours. And this is largely thanks to its enthusiastic adoption of smartphone contact tracking.
In early March, the Singapore government released its first contact-tracking app, called [3]TraceTogether . Developed as a joint venture between the Government Technology Agency (GovTech) and the health ministry, it works on a similar basis to the aforementioned idea, using Bluetooth beacons to identify contacts.
Another striking similarity is that it does not use location data, although this is largely by design. GPS is ineffective indoors and in built-up areas, and Singapore is one of the most densely populated nations in the world, with towering skyscrapers that play havoc with GPS signals.
Although downloading the app isn't mandatory, the health ministry can request access to TraceTogether data at any point. The government also says that the app will be retired, and cease to function, after the COVID-19 panic has subsided. It has also since open-sourced the project, [4]as we've previously noted .
TraceTogether represents the milder side of Singapore's anti-COVID efforts. The project cites over 1.1 million users. It already has more than 500,000 downloads on the Google Play Store. This is an impressive feat, given the population of Singapore is around 5.6 million.
Of course, the government of Singapore hasn't hesitated to use stronger measures, with potential fines of SGD$10,000 (around £5,700) for those who lie to the authorities about their movements. Offenders can also face up to six months in prison.
The app has also used the criminal justice system to enforce its lockdown, with one 71-year-old man arrested for refusing to go home after [5]being caught eating in public .
Israeli intelligence
Israel was quick to implement a lockdown. Beyond that, it's also taken measures above and beyond perhaps any other nation. On 17 March, the Cabinet approved emergency measures that would allow it to track infected individuals by their phone to inform those they may have had contact with.
What's noteworthy about this is how it leans on anti-terrorism technology previously designed to help with its nearly 70-year-old struggle with Palestinian and Arab Nationalists. It was never intended to be used against its own people.
It's also a rare example of a government using a top-down approach to contact tracing. Even Singapore, a country known for its authoritarian tendencies, hasn't made use of TraceTogether mandatory. Singapore banned chewing gum and once [6]caned an American teenager's derrière over a spot of petty vandalism.
Israel is well positioned to take advantage of surveillance technology in its battle against COVID-19. The country, surrounded by unfriendly neighbours like Syria and Lebanon, has long used intelligence to work against its enemies.
Early days in Europe
The UK is currently in week five of its national lockdown. For some countries, it's been longer. Others, shorter. But no matter where you go, the shattered sense of normality remains the same. Streets are quiet. The schools and bars are closed. The dole lines are swelling.
And with a vaccine still far away, the pressure is on these governments to find solutions that will allow a return to normal – or as normal as the situation can possibly allow.
Enter the Pan-European Privacy-Preserving Proximity Tracing: or PEPP-PT. This snappily named initiative, which is incorporated as a nonprofit in Switzerland, aims to [7]create a standardised method of tracking the interactions and movements of COVID-19 infected, which would form the basis of future apps.
Academics: We hate to ask, but could governments kindly refrain from building giant data-slurping, contact-tracing coronavirus monsters? [8]READ MORE
PEPP-PT is an independent initiative, although it claims to have won the support of seven national governments. The project, however, has [9]been fraught with turmoil , with many participants – including ETH Zurich, the CISPA Helmholtz Center for Information Security, and the Italian ISI foundation – withdrawing from the consortium over privacy concerns.
An [10]analysis of preliminary protocol design [PDF] by rival initiative DP-3T (Decentralized Privacy-Preserving Proximity-Tracing) highlighted several problems, both technical and operational.
The proposed implementation by PEPP-PT, called NTK, effectively creates a permanent identifier for each user, which can be linked to individual encounters. DP-3T argues this could allow for feature-creep, with the project potentially metamorphosing into "an instrument of surveillance with considerable human rights implications".
DP-3T also argues that this data could be used to track the daily activities and movements of individual users.
"In the extreme, the backend could assign special keys to certain users (note that a user has no way to know whether their key changes over time or not) and leak them selectively, enabling long-term tracking by third parties," it says. "We note that this also works for communities, as one could assign specific identifiers to a target group of people."
Ultimately, this dispute is reflective of the overall failure of Western governments to get ahead of COVID-19. Other countries, like Israel and Singapore, took decisive action, but many European nations hesitated when it came to instituting lockdowns. Though Singapore has had TraceTogether for over a month, there's no clear consensus on the continent when it comes to a contract-tracing system.
While Europe dithers, the exit from lockdown looks even more elusive. ®
Sponsored: [11]Forrester Build a Digital Experience Portfolio
[1] https://www.imdb.com/title/tt1598778/
[2] https://www.theregister.co.uk/2020/04/20/coronavirus_contact_tracing_academics_social_graph/
[3] https://www.tracetogether.gov.sg/
[4] https://www.theregister.co.uk/2020/03/26/singapore_tracetogether_coronavirus_encounter_tracing_app_lessons/
[5] https://www.channelnewsasia.com/news/singapore/covid-19-man-arrested-void-deck-eating-police-refused-leave-12620530
[6] https://www.nytimes.com/1994/03/16/world/singapore-journal-a-flogging-sentence-brings-a-cry-of-pain-in-us.html
[7] https://www.theregister.co.uk/2020/04/06/pan_european_privacy_preserving_proximity_tracing_plan/
[8] https://www.theregister.co.uk/2020/04/20/coronavirus_contact_tracing_academics_social_graph/
[9] https://www.theregister.co.uk/2020/04/20/pepp_pt_dp_3t_robert/
[10] https://github.com/DP-3T/documents/blob/master/Security%20analysis/PEPP-PT_%20Data%20Protection%20Architechture%20-%20Security%20and%20privacy%20analysis.pdf
[11] https://go.theregister.co.uk/tl/1936/-8554/forrester-build-a-digital-experience-portfolio?td=wptl1936
@KCIN
Would you mind to provide reliable sources for your statement?
The bulk of the verified evidence suggests that we don't know the answers at present but there's plenty of evidence released that suggests that we are doing enough research to probably know all of the answers within a year or two.
Until then it's all just posts on flakebook or twatter (oh darn it, auto-correct is not working).
Re auto-correct, lucky you! Mine is trimmed to change "a" into "ae" for the platform ending with ...book.
>Until then it's all just posts on flakebook or twatter (oh darn it, auto-correct is not working).
Seems to be working perfectly fine to my eyes. :-)
KCIN said
"It won't work"
You seem confident, yet something similar seems to have worked in Singapore, and that's with reportedly only 20% of their population using it.
Granted, they seem to have taken decisive action sooner than most other nations. I guess the conclusion is that pussyfooting around doesn't work.
Without commenting on your claimed facts, your logic is wonky. The objective it to tell you that you've been close to someone who's infective, not someone who's immune.
Oh great!
So our options are to be provided by either the worst Big Brother companies (Apple & Google), a "beneficial" dictatorship (Singapore) or Mossad?
And let me guess, these special apps will remain live long after the threat of COVID-19, just in case, much in the same way as all the draconian powers associated with the US Patriot Act are still with us, because terrorism.
No use
When mass transit resumes in London and other big cities, the number of possible contacts rapidly becomes too big to effectively track. (Just imagine one person changing trains at Oxford Street station - how many possible contacts?).
All the apps will be good for is "Big Brother" snooping.
The number of tests being carried out is also insufficient - if the government somehow manages to do 100,000 tests per day as the politicians are saying then it would take almost 2 years to test the population of the UK (over 66.5 million). By then the epidemic would already be over.
People can be infectious with covid-19 before showing any symptoms and many people never show symptoms but are still infectious.
The following paper from the ECDC - https://www.ecdc.europa.eu/sites/default/files/documents/RRA-seventh-update-Outbreak-of-coronavirus-disease-COVID-19.pdf - shows that people who never develop symptoms can still be infectious and also that the basic reproduction number (R0) for the virus is high at 3.28
Re: No use
Location data is not supposed to be harvested, in which case it is not Big Brother - except if you're in the UK which has, curiously, decided it wanted that.
If you can only tell who you've met and not where you met them, then you're doing tracing properly.
If that data gets permanently erased once the crises is over, it is not Big Brother but a serious medical initiative that is justified in these times.
Re: No use
A person walking through Oxford Street station to change from one tube line to another during the rush hour will typically be inside the infectious range (under 1 metre) of lots of people (possibly over 100). If the person is a non-symptomatic carrier then by the time that the first detected case appears the total of possible first and second stage contacts will total several thousands - too many to be of any use.
Distances are larger than they appear on globes
"One of the few coronavirus "success stories" has been Singapore. Despite its proximity to regional hot spots like South Korea and China"
They're not that close. Singapore to Seoul is over 4500 km. Even considering Hong Kong as China, it's still 2500 km to Singapore from there. Some comparisons might help. Singapore to Hong Kong is a similar distance as London to Istanbul or Vancouver to Chicago. Singapore to Seoul is a similar distance as London to Tehran or Vancouver to Guatemala City. Now it's true that people go between China, South Korea, and Singapore somewhat frequently. But they also frequently go to Australia, Japan, the U.S., India, and many other places. Some of these places have bad records and some have good ones. Proximity is not a good predictive measure of where cases will start to pop up--if it was, we wouldn't have expected Italy to have been one of those places.
Correction
Comment Most health experts agree stopping the coronavirus lockdown requires two things ...
No, it requires just one thing. That big business is satisfied after scavenging the rubble they created with this scam.
likely to be implemented here in oz, so how to manage it?
My immediate response is that the app will go onto a burner phone (actually my work phone - because I am not managing security on that, therefore it is not given data I personally wish to keep private)
Being a work phone, it is old and cheap, so may not support the app - but that is up to my employer to deal with.
Where I work carrying personal phones is actually grounds for termination - so a pretty strong argument there...
It would require my carrying the work phone whilst not at work - a bad thing, but many bad things are now normal, and fairly trivial as such things go.
Losing access to the work phone, through loss of job or whatever reason, would require buying a cheapo replacement, but not a problem.
Mobile phones are basically now cheap commodity tech, all the "flagship model" hype from apple/samsung is just marketing excrement.
Re: likely to be implemented here in oz, so how to manage it?
Network analysis is more sophisticated than what you think. Some known nodes can be used to identify unknown ones. By cross checking few repeated encounters with other data, like work colleagues identified via LinkedIn or other legal or illegal data eventually an unknown phone will be put in the right place in few days.
Singapore
One of the few coronavirus "success stories" has been Singapore.
Not a big deal since Singapore is a tropical country. Yes, I know, fake news outlets published reports stating that heath does not stop the spread of the epidemic. Unfortunately they presented the evidence with the contorted language typical of a lawyer, where the paper says that the virus in hot and dry conditions is less virulent although you cannot be sure 100% that it will be stopped they claimed that you can be sure 100% that it wouldn't work.
Why do Google and Apple develop new API?
Singapore has (and had for some time) perfectly working app doing exactly what G&A are planning, and it uses existing Android/iOS API. Strange things are going on.
It won't work.
There's now plenty of evidence that people who were antibody positive and no longer antibody positive