Schneier on Security

Clive Robinson • October 5, 2016 3:46 AM

In the article, Matthew Green who has looked at the Signal code sufficiently to form an oppinion says,

They would have to defeat both the encryption to and from the server and the end-to-end Signal encryption. That does not seem feasible at all, even with a Wi-Fi access point.

Whilst that is probably correct for attacking the algorithms, it might not be true for the practical implementation. That is the implementation on a device may have various communications time or power side channels that could be exploited.

Further as others above alude to it might involve some kind of malware injected into the device across the air that could get at device drivers etc for the keyboard and screen. However if that were the case you would expect the malware to be broader in scope and not just for WhatsApp.

Hence if I were placing the price of a pint on it I would be looking for some kind of side channel attack….

However it may be easier

Clive Robinson • October 5, 2016 5:40 AM

@ Roger,

Is a P2P messaging app like Bleep by Bittorrent Inc. safer and more secure?

Long answer short “no”.

Look at it from an attack developers perspective, to pay for the R&D of a product they need it to be marketable in a relativly short window of opportunity.

Thus they will find and develop attacks for all popular platforms and those that are likely to become popular. They will then try to “guess the market” to sell at an optimum time.

This means that likewise the larger agencies will have also developed attacks, but as they don’t sell product they do not have optimum market considerations to take into account, just getting the job done.

Thus it would be wise to assume that all messaging systems be they instant, Email, VoIP or other are insecure, especially if they are popular or standardised through the more traditional standards organisations.

What you need is a two part solution, a security device and a communications device. At it’s simplest the security device could be an old fashioned One Time Pad, the communications device anything that is currently convenient. What ever else you do never ever use a single device, and never put sensitive plaintext on the communications device no matter how urgent etc.

Provided you folow the rules you need not worry about the security of the communications device. However using as secure a communications device as you reasonably can, has the advantage of making any attackers job that not so little bit harder to isentify you as a person of interest.

Clive Robinson • October 5, 2016 9:12 AM

@ Here’s Johnny,

What about Bit Message is that any good?

I don’t know, because I trust none of them. Therefore I mitigate them by use of a secure device external to anything they might offer.

So I’ve no real need to check any of them from a “secure privacy” aspect.

That said it’s not just the “secure privacy” you need to wory about, but also “secure routing”, and this is an asspect that is a much harder problem to deal with.