Netsec – Telegram
Netsec
7.44K subscribers
22.4K links
This channel posts the feed from r/netsec.
For any suggestions dm @streaak
Donate to keep the bot running https://www.paypal.me/akhilgv
Download Telegram
Recent 1.4 billion password breach compilation as wordlist
http://ift.tt/2AVfJrx

Submitted December 19, 2017 at 09:25AM by stmiller
via reddit http://ift.tt/2yUqatE
Are EV certificates worth the paper they're written on?
http://ift.tt/2zOmP0p

Submitted December 19, 2017 at 02:04AM by 57696c6c
via reddit http://ift.tt/2BHDvuO
Google Advanced Security Not Actually Inconvenient
Since October of 2017 Google has been marketing it's "Advanced Protection Program" as a system that sacrifices ease of use for security. As if the average end user would be losing essential functionality or having to jump through serious hoops in order to use their account. In practice this is in no way the case- for people like ourselves AND the average end user. Because android phones and tablets keep you logged into Google even after a reboot of your device, you only need to use your nfc/bluetooth key one time. It's an obvious distinction for those of us who live and breathe data security but I think we could be explaining this to our less tech-savvy counterparts a lot better. Google also claims this feature is geared towards politicians and journalists but not necessarily our teenage children and financially vulnerable grandparents. It stinks that Google finally enabled such a basic functionality and then greatly damaged the possibility of it going mainstream.

Submitted December 19, 2017 at 06:02PM by sweepstor
via reddit http://ift.tt/2BHGgw5
Linkedin unread notifications count is open for everyone
http://ift.tt/2D3dDHx

Submitted December 19, 2017 at 07:41PM by RandomAdversary
via reddit http://ift.tt/2oLgBgB
GUI Tool for crafting ROP chains (WIP)
http://ift.tt/2D35Jhn

Submitted December 19, 2017 at 09:41PM by chillingswordfish
via reddit http://ift.tt/2D69liw
Thoughts/Advice
I'm currently going through a vetting exercise on a new vendor/solution for my business.The vendor is offshore, but is touting that they are compliant with a litany of things (NIST, HIPAA, PCI, ISO, etc) as well as saying that the employ a zero-trust policy with everything they do. That's all well and good. That should be a pretty basic standard when you're providing this kind of service.Here's where it gets interesting. As part of the review that my team performs, we ask how vulnerabilities are remediated. Here's the answer that we've gotten:"<software/application that they've written internally> maintains a clear separation between various systems. The policy is enforced realtime and on a continual basis. It is simply impossible to access any internal systems without first authenticating, authorizing role, and providing an intent. <Vendor> follows a zero trust policy for cybersecurity."This seems really off to me. First, I would never, in the history of ever, say that something was impossible to access. Further, the vendor has peppered throughout their entire vetting their tool and how it employs passphrase technology and no user IDs to make their solution "impossible to hack".Thoughts?

Submitted December 19, 2017 at 09:27PM by OldFennecFox
via reddit http://ift.tt/2klgLqw