Is the Coronavirus beginning of Mass Surveillance?

Nikola Basta
Arteos AI
Published in
6 min readMar 25, 2020

Imagine that one could know if cross paths with Coronavirus infected person. Would that be useful information for that person or his/her family? Imagine that there is an app that can share information about movement in a privacy-preserving way and let the officials track and tackle coronavirus hot spots.

Well, the future is now.

Safe Paths is the new app that offers a patient to, instead of relying on memory, provide their location data, where one has been and with whom one crossed the paths. Health officials can use this personal information from the location trail, and the app will notify users who came in close contact with diagnosed patients. Healthy user’s data never leaves their phone.

“This could help curb the spread of COVID-19”, says Ramesh Raskar at the MIT Media Lab, who Leads the team behind it.

Safe Paths is a free and open-source app. Developed by people at MIT and Harvard, as well as software engineers at companies such as Facebook and Uber, this application can find enormous applications in today’s Coronavirus pandemic.

How does it work?

Well, this app works by sharing location data between phones, which is fully encrypted in such a way that it does not go through a central authority. They have managed to get around the privacy concerns with the federated learning approach. In a nutshell, with this machine learning model, we can preserve privacy. Google first introduced it in 2017. Federated learning is also how Siri can be personalized without hoovering one’s data. Federated Learning uses collaborative learning, a share prediction model, and, at the same time, keep all the training data on the device. The device will download a current model, it will learn from the data on the cell phone and improve the model, and for the end, it will sum up all the changes as small model updates. Only this small update will be sent to the cloud, and it will be encrypted. It will instantly be averaged with all other user updates, which will improve the shared model. Essential to mention is that all the training data remains on users’ phones, which means that individual updates are not stored in the cloud.

https://ai.googleblog.com/2017/04/federated-learning-collaborative.html

So let us get back on SafePaths. It will use this model to preserve privacy and inform users if someone with Coronavirus passes by or if the user has been in contact with an infected person. It is important to say:

  1. That person has to share the data using the app.
  2. You would not know who is that person. You just know that you have contact with, or pass by an infected person.

Apart from this type of sharing, a person who is positive to Coronavirus can choose to share location data with the health officials also. After that, they can make it public if there is a mutual agreement.

With this approach, only specific locations would be closed and subject to disinfection, which is far better than general shutdown and quarantine for the whole nation and world.

Will it make a real difference?

This application can make a difference but only with a global scale, or to translate it, if enough people use it. South Korea already implemented this approach and has started to use people smartphones to watch quarantined patience. This approach of isolating the coronavirus hot spots proved to be effective.

On the other hand, incomplete information can lead to a false sense of security. For example, the application can tell which places are secure, but the thing is, it will only know where the infected person has been, not where the virus will be in the future.

What about privacy?

Aggressive measures were suggested by The World Health Organization in this biological war against the Coronavirus. Apart from identifying individuals who are infected and properly isolating them, one of the most important measures is also to identify the people they have been in contact with. This way, those people can also be tested and isolated as well as the place where there are located. We saw that the app mentioned above properly tackles privacy but have in mind that all our phones can be tracked even without this app. This topic of surveillance from the government's perspective would be hard to justify in democratic countries. We would need something drastic to happen so that this becomes normal and acceptable - something like the State of Emergency.

Israel Case

So, with all of this been said, let’s take a look at how it can be, or to be more precise, is implemented in reality. Prime Minister Benjamin Netanyahu recently accounted that his country will start with the use of advanced digital monitoring of infected patients, which instantly raised a lot of questions and privacy concerns and accusations of mass surveillance.

This technology will primarily rely on smartphones and already been used in counterterrorism operations. In the past, it was not used against the Israeli citizens apart from those accused of a crime. So technology is there, it has already been used, and now it is just a matter of political decision whether it is going to be used again with a broader spread of surveillance power.

So now that this dramatic course of action has been confirmed, it was also announced that these tools would not be used to enforce quarantines but rather to help authorities track the path of confirmed patience on COVID-19 in order to identify people with whom they have been in contact with.

In a nutshell, the technology is used by Israeli security services to track the location of people in real-time as well as post festum. This technology allows them to retrace steps, locations, and people with whom they have been. We could assume that this could not be spread that easily to the whole nation without resistance from another side. Attorneys said that the number of restrictions and rules needs to be applied and put in place so that the government would not violate the privacy of individuals.

The first thing was that approval hast to be received from the people, more precisely, infected patience but also the rest of the possible infected citizens. This process could take time, and the time is maybe the most important thing when we speak about Coronavirus. That is why we now have the State of Emergency where normal democratic rights are on the test. Moreover, something like this could not take place and be deployed without further discussion, mainly parliamentary and judicial.

A media and technology researcher, Tehilla Shwartz, criticized this plan also.

“A state of emergency doesn’t mean it’s acceptable to turn the State of Israel into a police state,” she wrote in a series of tweets. “Infringing upon rights is easy, returning the situation to what it was before is much harder.”

It is fair to say that Israel is not the only country that reaches out to these measures. These measures should be subject to big discussion before becomes acceptable in the broader audience with all pros and cons. We saw a responsible approach used in the Safe Paths but also political overnight decisions that could and will harm the process of trust-building between people and authorities, which is crucial for utilizing the full power of advanced machine learning solutions like Federative learning.

--

--

Nikola Basta
Arteos AI

Optimizing business processes, minimizing costs and maximizing profit using machine learning and deep learning solutions.