News for hackers. Stuff that matters.
Re:There is no WhatsApp 'backdoor' by edejong on Friday January 13, 2017 @07:10PM (#13395940)
I'd say the risk of replay attack is not proportional to the risk of authenticating and authorizing on a fully compromised machine. It's comparing a candle light with a blazing fire.

A definition issue with regard to "token": I sincerely do not think a device with display and keyboard used to sign transactions can be called "token". I'd say something like: trusted signing processor. But then again, I'm not a security specialist.

Re:There is no WhatsApp 'backdoor' by akvadrako on Saturday January 14, 2017 @06:18AM (#13397944)
Well the big issue is that it prevents future access to your account. Let's say you have a simple 2FA device (no screen) and are using online banking. First you login via a compromised machine. The attacker MITM'ed you, so can see your account.

1. The bank should require a confirmation with your token to send money. If you don't send anything, the attacker can't either.

2. In the future, after the logout timeout, you know the attacker can't even read your account.

It greatly reduces the attack surface you need to worry about. Any attack they do must be right then.