Airnb and the cell provides are good examples [1],[2].
The cell provider location data is the most insidious. They add noise to it, but the central limit theorem is a real thing and people who buy the data are aware of that.
I’m aware of what the central limit theorem is, but I’m confused as to what you mean here. Do you mind clarifying? It sounds like you’re saying the CLT leads to deanonymization. I don’t see how the two are related.
Take several measurements and average them. The variance scales like 1/n_samples, which can take you from city-block resolution to building resolution quite rapidly.
There are conditions for the central limit theorem to hold though. One can generate random noise which violates them. It's certainly hard to guarantee privacy, but it's not trivially easy to hack if they're halfway smart about it.
Correct. This "if the product is free, then you're the product!" thinking needs to go away. The reality is numerous companies sell your data today, regardless if you're paying for the service or not.
It is possible for two things to be true at the same time. You are right, there are paid apps that sell location data to brokers... but the vast majority are free apps that rely on mass monetization of users for their income.
The cell provider location data is the most insidious. They add noise to it, but the central limit theorem is a real thing and people who buy the data are aware of that.
[1] https://www.nytimes.com/2020/01/15/technology/data-privacy-l...
[2] https://www.vice.com/en_us/article/nepxbz/i-gave-a-bounty-hu...