Don't Do KRACK - Breaking WPA2 WiFi Protocol

If this is a simple case of masquerading, then once the hacker gets the security key from the client, wouldn't the hacker have what it needs to compromise the network security? Or despite this exploit, is the key still secure and only (client) data compromised?

If you're using WPA AES, it's only the client. You can't join the network.

https://arstechnica.com/information...ck-attack-destroys-nearly-all-wi-fi-security/

But other modes are more vulnerable.

I presume that most people here are using AES, so even on unpatched devices this is not too bad. And most stuff you're doing online is already encrypted, so this just removed the second layer of encryption. But if you're paranoid, just turn off WIFI when you're buying something or depositing a check :D
 
Last edited:
My understanding was that KRACK only exposed data back and forth between an unpatched device and an AP, NOT that this exploit can lead to stealing the key and compromising the network.

Are you now saying that this exploit can lead to theft of the key and allowing unauthorized devices to access the network?

It is basically a MITM attack. What all it can get varies depending on AP and client. All compromise the network in at least one direction and some compromise it in both directions.
 
This attack is done via NONCE reuse.

When a client joins a network, it executes the 4-way handshake to negotiate a fresh encryption key. It will install this key after receiving message 3 of the 4-way handshake. Once the key is installed, it will be used to encrypt normal data frames using an encryption protocol. However, because messages may be lost or dropped, the Access Point (AP) will retransmit message 3 if it did not receive an appropriate response as acknowledgment. As a result, the client may receive message 3 multiple times. Each time it receives this message, it will reinstall the same encryption key, and thereby reset the incremental transmit packet number (nonce) and receive replay counter used by the encryption protocol. We show that an attacker can force these nonce resets by collecting and replaying retransmissions of message 3 of the 4-way handshake. By forcing nonce reuse in this manner, the encryption protocol can be attacked, e.g., packets can be replayed, decrypted, and/or forged. The same technique can also be used to attack the group key, PeerKey, TDLS, and fast BSS transition handshake.

Implementations can be patched in a backwards-compatible manner. This means a patched client can still communicate with an unpatched access point (AP), and vice versa. In other words, a patched client or access point sends exactly the same handshake messages as before, and at exactly the same moment in time. However, the security updates will assure a key is only installed once, preventing our attack. So again, update all your devices once security updates are available. Finally, although an unpatched client can still connect to a patched AP, and vice versa, both the client and AP must be patched to defend against all attacks!
 
My understanding was that KRACK only exposed data back and forth between an unpatched device and an AP, NOT that this exploit can lead to stealing the key and compromising the network.

Are you now saying that this exploit can lead to theft of the key and allowing unauthorized devices to access the network?

It's not theft of a key. It's forcing the application to reinstall a crafted key.
 
Finally, although an unpatched client can still connect to a patched AP, and vice versa, both the client and AP must be patched to defend against all attacks!

This is interesting, and the first I have heard of it. My Unifi AP's received a patch yesterday, but according to the release notes they were only patches for their STA functionality, when they act like a client in a bridge functionality.

This implies they still need a patch for standard AP mode.
 
That's a good summary, Schtask. I made a thread with some info and instructions to help patch up based on mine own experience thus far. https://hardforum.com/threads/krack-wpa-vulnerability-patch-everything.1946100/

Ugh, it feels like BlueBorne was just yesterday...

Could be worse. Like Infineon having an weak RSA key generator, which makes electronic ID cards in few countries vulnerable. Like in Estonia or Slovakia. You know, it is used for such irrelevant tasks like tax, VAT returns, communicating with courts etc.
 
Here is the best info I found inside the Unifi thread for their patch:


waterside said:
All but one vulnerabiliy in the set collectively called 'KRACK' attack WiFi stations (aka 'clients'). There is nothing that can be done on the AP side to change this, prevent this, or otherwise provide any mitigation for this. The only solution is to update each and every client.



The important point: If you have a vulnerable client device it is equally vulnerable regardless of whether or not the AP to which it is associated has been "patched".



The reason that there are recommendations and statemtents to patch both sides is that:

There is one vulnerability against 802.11r ("fast roaming") that may impact some APs Some APs also include station ("client") functionality and so this specific feature of the AP would still be vulnerable itself.
UBNT has released updated firmware to address #2 and have confirmed that #1 is yet to be remediated but recommends this be left disabled as mitigation, which is its default state.
 
Back
Top