WPA2 encryption security update for older phones?


#1

I’m using an older phone, a 1st Gen Moto X, that doesn’t get AndroidOS updates anymore. Will I be getting the security patch for the WPA2 exploit that was recently discovered? Has it been issued yet?

I’m referring to this:


Moto X Pure KRACK fix
#2

I am also interested in whether RW/Moto will address this. This is the type of severe security bug that, for vendors who are reasonably security-conscious, elicits a patch even for EOL products. For example, a year ago, patches for the QuadRooter vulns were released for Moto X 1st gen.


#3

Hi @michaels.hg1x8x,

It would be up to Motorola not Republic. My gut instinct is no. More here: Where's My Android Update? Understanding the Update Process.

It has not. Google is promising to address the vulnerability with November’s security patch, however, individual manufacturer’s would still need to push that security update to their phones.

It’s important to note, this vulberability can also be addressed by router manufacturers with firmware updates. In other words, if the router one is connected to is patched, it does not matter whether attached devices are.


#4

Hi @michaels.hg1x8x and @greg.b,

The decision to update any phone is out of our hands. You’ll want to check with the phone’s manufacturer. See: Where’s My Android Update? Understanding the Update Process for more details about how updates are released.


#5

I agree with @rolandh gut and the Legacy phones will not see this
as far as Lenovo/Motorola is concern all Legacy phones at at end of Software life and will not see any more updates
the TL:DR below
Legacy phone with Custom ROM would also need Republic to test their added code in a new update and from pass experience it would take 3-9 months to get an update (and note Republic OS build teams have been moved on to other job functions as Republic no longer needs a custom ROM


#6

AFAIK this is not true, though there is still misinformation floating around in these early hours of this release, so I may be proven wrong. Even if the router is patched, the client can still be attacked. This is partly why this vulnerability is such a big deal – everything needs to be patched. In fact, although the vulnerabilities do affect routers (running in AP mode), the typical attacks affect WiFi devices in station (i.e. client) mode.


#7

While this is fully true for 3.0 phones, it is only partly true for legacy phones – if Moto releases an update, RW still has to process it and push it out to RW devices. Yes, it does depend on Moto initially, but RW still has to play an important part (that requires some investment) in getting the update to its customers.


#8

As I read it, the primary attack vector is through the retransmission of cryptographic keys. Therefore, either the router could be patched not to retransmit those keys or the client could be patched not to allow reuse. More from the folks who discovered the vulnerability here: https://www.krackattacks.com

The above said, the myriad of router configurations and with no clear way to tell if one is connected to a patched router or not (if not under one’s control), make patching clients more practically effective.


#9

@rolandh – the authors provided a POC of just one method of attacking the vulnerable products – there are some ten different CVE’s associated with this, which have to do with key reinstallation, not key retransmission.

As I understand it, the typical exploit here involves MITM’ing the WiFi connection, which means there are some reliability dependencies on the physical positions of the AP, victim, and attacker, but the MITM nature of this is part of why the client needs to be patched too.


#10

Candidly, the practical effects of most security vulnerabilities as reported by the press (in particular the mainstream press) are overstated. Currently, proof of concept is all we have. The existence of multiple CVEs do not necessarily mean practical attack vectors exist in the wild or ever will.


#11

Everything I have read is if the router is patched than ever device on the network is safe. I read the link that @rolandh has posted here on this subject.


#12

So here is the brass tacks of the exploit. One it is just at the proof of concept phase. No one in the wild is known to even have used it. As @rolandh has stated, the press likes to overblow these issues. Second, the carriers, whether it be Republic wireless or any other carrier is not going to be the party to patch this issue. The manufacturer of WiFi devices (phones, routers, etc.) or software developer on Computers will be the party that will patch your device or not. Microsoft has already patched it in Windows 10 from what I have read.

Second, if your device is over two years old, it probably won’t be able to see any update.

Finally, any https level encryption will not be vulnerable to my understanding. If you run a VPN on device, it should not be vulnerable. Running a VPN on a RW device will cause call quality issues on WiFI because of the additional latency.


#13

@coreyk, @rolandh - I read the researcher’s writeup earlier this morning, and here’s the important quote: “Our main attack is against the 4-way handshake, and does not exploit access points, but instead targets clients.” The typical attack involves an RF MITM.

Sure, HTTPS is a good thing, but there is a very good reason why WPA2 exists. For one example, even with HTTPS you can still observe what site the victim is connecting to (via DNS requests, IP’s, and/or the cert in the TLS handshake).


#14

You might be able to observe the site, but you get no personal data.


#15

@coreyk - I suppose that depends on what you consider to be “personal data” that’s worth protecting; however, even the concept of defense in depth (as illustrated by the myriad attacks against TLS over the years) means that this needs to be patched.


#16

The chances for getting an Android security update for the Moto X 1st gen are slim to none, and I would lean heavily toward the none.


#17

And device manufacturers will be the ones that will do it or not. Asking Motorola to patch a 4-5 year old phone would be like asking Ford to have a recall on a 1980s Ford Mustang for a failure today. I know that isn’t the answer that anyone would want to hear, but it is the reality of the situation. This is if your assertion is true that a client device on a network where the router has been patched is still vulnerable and I at this point do not believe that to be true.


#18

Y’all may very well be correct on this… though I’d suspect this would be more likely to be on Moto’s radar screen than a root hole would.

I’ve looked into this further, and I’m still convinced that patching the AP or router won’t protect the client against the 4-way transient key negotiation attack. Looking over the source code for the patches to a rather popular AP/router software product¹ (which is, I believe, also used in our phones for wifi tethering), they are not patching it to prevent retransmissions of Msg3 of the 4-way handshake (which is what triggers the client to reinstall the key). Instead, they are patching the AP/router software to deal with the fast transition roaming side of this flaw and so forth.

So, does patching the router typically protect the client? Yes, but only from exposure to attacks against the AP/router itself (e.g. 802.11r fast transition roaming); it does not protect from the most impactful version of the attack that has been demonstrated.

¹ hostapd - see http://w1.fi/security/2017-1/


#19

I agree that the main problem is being reported as being in WiFi clients rather than APs; more details including excerpts from the paper at Ars Technica. The tables shown in that article only mention Android 6.0 and later as being affected, but it is likely that earlier versions such as Android 5.1 (which is what most of us Gen-1 Moto-X users have) are probably still vulnerable and most likely will remain so forever.

That means that as users we should assume that all WiFi connections can be snooped, as they could be whenever we connect to the Internet over an open, public WiFi. Many websites nowadays use secure connections so the content of communications to such sites are still secure, but the fact that we are connecting to a particular server might not be.

Can anyone confirm that all communications between my phone and RW for services such as SMS, Phone calls and account administration travel over secured links? That’s the main thing I’m worried about at this point, and I would be surprised if any of those connections were not over an encrypted channel.


#20

Phone calls are not secured, I believe SMS are secured between your phone and the Republic servers, but it has been awhile since I have seen that question so I am not 100% positive on that one.