3 ms·
But it isn't a grey area: it simply doesn't work. It doesn't matter if it correctly identifies most people: it has to correctly identity most terrorists, and it
by saurik 2y ago
But it isn't a grey area: it simply doesn't work. It doesn't matter if it correctly identifies most people: it has to correctly identity most terrorists, and it simply doesn't do that, because if you are a terrorist you just keep rotating through IP addresses on cloud providers and VPNs until the entire service is burnt. It isn't that it sometimes doesn't work: it's that it doesn't work at all when it actually needs to work. We could argue that the services shouldn't let you do that in the first place, but the reality is that services currently do work like that, no one is trying to change that, and if they did try to change it we would all be even less happy with the resulting even-more-powerful surveillance state.
- nradov 2y agoWrong. At this level there is no compliance requirement to specifically identify "terrorists". And the sanctions against Iran, while partly based on state sponsorship of terrorism, aren't limited to just designated terrorist entities. Most services aren't required to blanket block all traffic from Iran. Only certain specific transactions are prohibited. But a lot of companies choose to block everything identified as coming from Iran (and other sanctioned countries) just to play it safe.