Our Arguments Against the French Contact-Tracing App StopCovid

Posted on


Last week, French President Emmanuel Macron invited the French Parliament to discuss the potential use of StopCovid, the contact-tracing application that his government is developing. We just sent out to the Members of the Parliament the following summary of arguments regarding this application.

The StopCovid application would be useless, could endanger our civil liberties and could even worsen the health crisis.The French administration and Parliament must stop investing human or economic resources in this vain and dangerous project. The real emergency lies everywhere but here.

Dubious Effectiveness

Low Adoption Rate

  • According to first approximations, at least 60%, but more probably 80% to 100% of the population would need to use the app in order for it to have a significant effect, and this assuming the app produces reliable data ;
  • Only 77% of the French population owns a smartphone, and this number drops down to 44% for those over 70 years old, whereas they are among the most vulnerable;
  • Many people do not know how to turn on Bluetooth, and many refuse to have it always turned on for practical reasons (saving battery) or security reasons1Many of the smartphones in use are not equipped with the latest security updates, and flaws in the Bluetooth protocol have been discovered over the last few years.;
  • In Singapore, around 16% of the population used a similar app. This didn’t prevent an eventual lockdown.

Imprecise Rresults

  • it is unlikely that tests will be made as readily available as they would need to be, preventing people from reliably reporting their true condition. Relying on people self-diagnosing runs the risk of producing very high numbers of false-positives.
  • There seems to be no consensus as to the duration and the distance of contact that justifies warning a person who came “in “contact” with another contaminated person.
  • In some very densely populated areas (some neighborhoods, shopping malls, large companies), one would see an explosion of false positives, which would effectively render the app useless.
  • Blueetooth range seems to vary widely from a device to another, and precision is not necessarily good enough to provide reliable results2See the’analysis of ACLU : “Other open questions include whether Bluetooth is precise enough to distinguish close contacts given that its range, while typically around 10 meters, can in theory reach up to 400 meters, and that its signal strength varies widely by chips et, battery, and antenna design” .

Counter-Productive for Public Health

  • By creating a false sense of security regarding public health, the app could lead people to reduce the basic protective behaviour, while failing to reliably notify people.
  • The energy and costs required to develop the app would not be invested in more effective solutions such as producing more masks, ramping up testing or promoting basic protective behaviour.
  • Implementing surveillance systems would increase the already strong feelings of defiance in part of the population towards the State. Without confidence in the system, potentially ill people could be incited to hide their symptoms from medical services fear of negative consequences.

Liberties Sacrificed in Vain

Discrimination

  • Whether by making it mandatory or through excessive social pressure, people deciding not to use the app could risk losing their jobs or be denied access to public spaces (cf. case in Italy (Italian: https://www.laprovinciacr.it/news/italia-e-mondo/246504/coronavirus-ferrari-app-tracciamento-per-dipendenti.html)), making their consent non-free and thus null and void.
  • One of the worst possibilities of discrimination would be to facilitate access to testing to those who use the application.

Surveillance

  • If the app were to be used by part of the population, one can fear that the government would then be able to more easily impose its usage to the remainder of the population, against its will. It is noteworthy that none of the securitian and freedom-destroying measures taken in times of “urgency” are ever revoked: this ratchet effect plays strongly in the justified defiance towards these control measures.
  • The goal of the app (warning targeted people) is by essence incompatible with the legal concept of anonymity. At best, it is pseudonymity, which does not protect against the risk of individual surveillance.
  • Publically releasing the application’s code under a free and open source license as well as using reproducible compilation methods would be the mininum requirements against abuse, but non-sufficient in and of itself.

Normalisation of Surveillance

  • No one can predict how long the app would be deployed for.
  • Once deployed, it will be easier for the government to add coercive funcitonality to the app (control and enforcement of quarantine at the individual level).
  • The app pushes us to submit our bodies to constant surveillance, which gradually reinforces the social acceptability of other technologies, such as facial recognition, or automatic videosurveillance, which are currently largely rejected.
  • Technological solutionism: the app reinforces the blind belief that technology and surveillance can be the main answers to health, environmental or economic crises while on the contrary, they draw attention away from the solutions: scientific research, adequate financing of public services…

Deploying an app whose objectives, technology and usage carry significant risks for our society and our freedoms, for likely mediocre results (possibly even counter-productive ones) is not something we can consider acceptable – nor is it for many French people. The media and political time, but also the budget allocated to this, would be better spent on informing and protecting people (and health workers) with proven methods such as providing more safety masks, medical equipment and blood tests.

References

1 Many of the smartphones in use are not equipped with the latest security updates, and flaws in the Bluetooth protocol have been discovered over the last few years.
2 See the’analysis of ACLU : “Other open questions include whether Bluetooth is precise enough to distinguish close contacts given that its range, while typically around 10 meters, can in theory reach up to 400 meters, and that its signal strength varies widely by chips et, battery, and antenna design”