|
|
|
|
|
by tga_d
1146 days ago
|
|
I'm curious why you seem to think that Tor is more legitimate to block than those behind CGNAT. There's been plenty of research showing on a per-connection basis, Tor is no more prone to malicious activity than connections from random IPs, and that it's only on a per-IP basis malicious activity is more likely. I.e., it's the same phenomenon as why CGNAT causes collateral damage. You could argue that Tor is opt-in and therefore less worthy of protection, but saying "users who want extra privacy deserve to be blocked, even when we know (as much as one can know) that they're not using it for malicious reasons" seems like a fairly dystopian premise. I'm actually kind of glad more people are becoming aware of this problem, and hope it finally spurs more interest in mechanisms that divorce network identity from IP addresses -- including the work Cloudflare is doing on Privacy Pass! |
|
However, the sad fact is that Tor is abused for a LOT of malicious traffic, much more so than any VPN provider, let alone normal ISPs using CGNAT. The anonymity combined with its free nature make it very attractive for bad people to use Tor for bad things without any reasonable fear of getting caught.
An outright block for Tor traffic is definitely out of the question, but adding CAPTCHAs to sensitive things (like account signups, expensive queries, etc.) is sadly a requirement these days.
Blocking exit nodes does nothing to protect your website's security, but it sure as hell cleans up the logs and false positives in your security logs. It's not just Tor, though, there are also some home ISP networks that don't seem to care about the botnets operating inside their network.