System Level Protection ??

We are now hearing lots of news about attacks in top notch companies like FB, LinkedIn, Twitter etc. The main reason stated is vulnerability in Java plugin which makes the attackers to penetrate.

Though Oracle has released a patch, I'm not sure whether it's a temporary patch or a permanent fix for the problem. Last month I came across one colloquium by Dr.Michael Franz ( http://www.ics.uci.edu/~franz/ ) UC Irvine held at UT Dallas. The talk was about "Software Immunity via Large Scale Diversification".

The abstract is below for reference.

Abstract

-----------

We have been investigating compiler-generated software diversity as a defense mechanism against software attacks. Imagine an "App Store" containing a diversification engine (a "multicompiler") that automatically generates a unique version of every program for every user. All the different versions of the same program behave in exactly the same way from the perspective of the end-user, but they implement their functionality in subtly different ways. As a result, any specific attack will succeed only on a small fraction of targets. An attacker would require a large number of different attacks and would have no way of knowing a priori which specific attack will succeed on which specific target. Equally importantly, this approach makes it much more difficult for an attacker to generate attack vectors by way of reverse engineering of security patches.

We have built such a multicompiler which is now available as a prototype. We can diversify large software distributions such as the Chromium web browser or a complete Linux distribution. I will present some preliminary benchmarks and will also address some practical issues such as the problem of reporting errors when every binary is unique, and updating of diversified software. 

-------------

So prof and his team had built a multicompiler which would build unique version based on a seed value ( like how most cryptographic algorithms work) for each and every user in the world. All the different versions built behave in exactly the same way. So the end user wouldn't be able to distinguish it. From the attacker perspective, this would require a large number of attacks and he/she would have no way of knowing the seed value through which the executable is made. Even if they can guess the seed, the attack would be minimal restricted to a group of people or single user.

This process makes the life difficult when a patch is released as it needs to be catered to each and every user as the company issuing patch should keep track off the versions running in every user, retrieve the seed used in the executable and generate the new executable based on some new seed/existing seed. In the talk professor had mentioned this new technique is currently deployed at Mozilla.

 Having explained the idea, would this really work??

This means by doing some nifty things with the executable can we really increase the security? 

Prof and his team would have definitely thought on various aspects.

Lets observe as time progresses..

What do you guys think on this?