> It's controversial because some people believe, reasonably, that opt-in is the only ethical model for collecting potentially sensitive information like browsi

> It's controversial because some people believe, reasonably, that opt-in is the only ethical model for collecting potentially sensitive information like browsing history, "anonymized" or not. You presuppose that it's reasonable, but is it? If you can prove that you're doing everything right to protect the user, should you not be allowed to collect the data? If we just call every non-opt-in data collection unethical, no matter if it can actually negatively affect users or not, then you're levelling the playing field in favor of those that actually collect and use data unethically. Because you're essentially saying that it's all the same, unless you're not collecting data at all. And for some services, you simply do need to collect data, and sometimes you even do need it to be non-opt-in, because otherwise your data sample will have a severe bias. > Mozilla stated "Cliqz uses several techniques to attempt to remove sensitive information from this browsing data before it is sent from Firefox"[1], which is very different from "no personal data will be collected". Even just a list of domains can be enough to reveal someone's identity. If you just so happen to read one sentence further than that, it says "Cliqz does not build browsing profiles for individual users and discards the user’s IP address once the data is collected.". That says exactly that no personal data is collected. And yes, a list of visited domains can be used to identify a person, but it's not necessary to store the visited domains aggregated by each individual user for this. They'll just intermix the domain lists from all users.