I've had many replies from my #StopCovid tweet from yesterday.
Let's have a look at the latest announcements from @platombe as he details *officially* my main issue.
Let's have a look at the latest announcements from @platombe as he details *officially* my main issue.
The talk is available at and unfortunately begins with some sophism like "if it's not efficient it can't be dangerous", but let's move to the important part.
He then states the application is OSS. This is true, but this doesn't solve anything at all.
1/ many companies use OSS. Palantir, Facebook, Google use OSS. That doesn't guarantee your data's privacy.
2/ The StopCovid data is stored in a MongoDB database, which isn't OSS.
1/ many companies use OSS. Palantir, Facebook, Google use OSS. That doesn't guarantee your data's privacy.
2/ The StopCovid data is stored in a MongoDB database, which isn't OSS.
Then there's the quick, but most important part: the data will be used for statistical analysis.
That's it. The data isn't encrypted, or secured in the "republic's most secure file" like @cedric_o said.
It's even made to be shared.
That's it. The data isn't encrypted, or secured in the "republic's most secure file" like @cedric_o said.
It's even made to be shared.
I've worked on "anonymized" social security data that was shared with one of France's biggest insurance company. You'd be surprised how easy it is to find people from that dataset.
Cambridge Analytica also had "anonymized" data, we all know how this ended.
Cambridge Analytica also had "anonymized" data, we all know how this ended.
I just don't believe centralizing all of the country's contacts in one database, and then sharing it, is a good idea. And as this is managed by people who don't seem to understand that very well, or are not very honest, this worries me even more.
