The logic behind the measure, which experts call ‘timing analysis’: The more nodes in the Tor network are surveilled by the authorities, the more likely it is that a user will attempt to disguise their connection via one of the monitored nodes. By timing individual data packets, anonymised connections can be traced back to the Tor user, even though data connections in the Tor network are encrypted multiple times.

So if you already have a suspect, you can see if you're probably wrong. I think this kind of attack has always been possible. There has, for many years, been talk of adding random traffic to confound this kind of thing, but it would come at the cost of performance.

True. de-anon for a specific user was already possible. Also, if you are already nation state target, there are other fields involved and Tor is just one of them. Correlation between other fields (real life social graph, mobile phone, etc..) might be much easier.

Tor as I2P are efficient, but blindly using them in any sort won't make them efficient to you.

https://blog.torproject.org/tor-is-still-safe/

Apparently, CCC has info they haven't yet shared with Tor. I'd like to know why. Does CCC want Tor to be vulnerable? That would certainly surprise me.

Is Tor still safe to use? | Tor Project

This blog post is a response to an investigative news report about a large-scale law-enforcement attack that managed to de-anonymize a user of an old version of the long-retired app Ricochet.

^{blog.torproject.org}