Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

UserVoice Security Incident Notification

community.uservoice.com

Post reply on HN
Print view View original

1–10 of 14 posts

Re: UserVoice Security Incident Notification

#2

Post by RossP »

"In late April, the UserVoice security team learned that an unauthorized party illegally accessed one of UserVoice’s backend reporting systems and was able to view user data on a small subset of users. The user data includes name, email, and a hashed password and salt. Unfortunately, the passwords were hashed with the SHA1 hashing algorithm, which by today’s standards is considered weak. As such, we’re resetting the passwords for all users in our database."

Further information: https://status.uservoice.com/incidents/fb7ml8b3nphf

Re: UserVoice Security Incident Notification

#5

Post by snoonan »

nsgf wrote: post #4
nacs wrote: post #3

Just got an email from Uservoice about this. Apparently I'm part of the "0.001%" that was affected in the breach.

Me too. Maybe 0.001% is not accurate.

seconded. I got one too. It seems unlikely we'd converge here if it was only a tiny fraction of users...

Re: UserVoice Security Incident Notification

#6

Post by nsgf »

snoonan wrote: post #5
nsgf wrote: post #4

Earlier quoted context omitted.

Me too. Maybe 0.001% is not accurate.

seconded. I got one too. It seems unlikely we'd converge here if it was only a tiny fraction of users...

Funny thing, 37 minutes after receiving the first one, i received another notification on a secondary e-mail (also a gmail account) that is used for a toy project's free UserVoice account and is totally unrelated to the first one).

Re: UserVoice Security Incident Notification

#7

Post by tempestn »

Another thread on the incident report here: https://news.ycombinator.com/item?id=11664713 https://status.uservoice.com/incidents/fb7ml8b3nphf

There's a bit more info in this one about exactly what was compromised though. While I can understand the abundance of caution in resetting passwords despite only hashes and salts being lost, it is odd that they would "[presume] the attackers may be able to decrypt the passwords," assuming they're using strong encryption.

Re: UserVoice Security Incident Notification

#8

Post by runesoerensen »

tempestn wrote: post #7

Another thread on the incident report here: https://news.ycombinator.com/item?id=11664713 https://status.uservoice.com/incidents/fb7ml8b3nphf There's a bit more info in this one about exactly what was compromised though. While I can understand the abundance of caution in resetting passwords despite only hashes and salts being lost, it is odd that they would "[presume] the attackers may be able to decrypt the password…

I wouldn't call resetting passwords an "abundance of caution" in this case. It's very likely that the attackers are able to retrieve passwords when they have the SHA1 hash and the salt (not exactly by decrypting though).

Here's a good blog post how and why this is problematic: https://www.troyhunt.com/our-password-hashing-has-no-clothes...

Re: UserVoice Security Incident Notification

#9

Post by tempestn »

runesoerensen wrote: post #8
tempestn wrote: post #7

Another thread on the incident report here: https://news.ycombinator.com/item?id=11664713 https://status.uservoice.com/incidents/fb7ml8b3nphf There's a bit more info in this one about exactly what was compromised though. While I can understand the abundance of caution in resetting passwords despite only hashes and salts being lost, it is odd that they would "[presume] the attackers may be able to decrypt the password…

I wouldn't call resetting passwords an "abundance of caution" in this case. It's very likely that the attackers are able to retrieve passwords when they have the SHA1 hash and the salt (not exactly by decrypting though). Here's a good blog post how and why this is problematic: https://www.troyhunt.com/our-password-hashing-has-no-clothes...

Do they say somewhere that they're only using sha1 though? That's sort of what I meant: if bcrypt or scrypt is used, with an appropriate work factor, the risk should be very minimal. The fact that they're assuming it's not suggests they are using weaker encryption.
Post reply on HN