2 ms·
The 'email does not exist' decision they make here is debunked for one single reason: risk assessment. Is there a risk associated with implementing this use cas
by peterwwillis 12y ago
The 'email does not exist' decision they make here is debunked for one single reason: risk assessment. Is there a risk associated with implementing this use case that outweighs not implementing this use case? Here the use case is 'tell someone if their email is [not] registered', and the risk is 'identifying existing user accounts at login time'. So let's see which is appropriate.
In the best case, implementing this feature lets a user know which of their e-mail addresses are registered on your site, giving them an easier log-in. In the worst case, an attacker can immediately identify what users have accounts on this site, exposing the users to a violation of privacy.
Some counter that site registration will always provide this information when attempting to register an existing e-mail address. But this is easily mitigated: a captcha as well as holding off verification until the last step of registration puts the brakes on the attacker, making it more difficult to scan through potential accounts. By slowing down their attack methods you're helping to reduce risk.
And finally, do we really need to tell the user what e-mail they used? I'll wager that the great majority of users enable browser form-filling, and so their e-mail is sitting right there in the login box waiting for a password. And i'd also wager that most users with multiple e-mails will be cognizant of which they use for an account on a site like yours. So to avoid the few times a user doesn't save form fields and can't remember which e-mail they use for sites like yours, bad people get an easier way to do bad things.
For these reasons I think it's clear that the minor inconvenience in a very small number of cases is worth it to help slow down privacy violation en-masse.
---
Nit-pick: why would God call the field 'User' if they really meant 'Email'?