Hillary Clinton wants “Manhattan-like project” to break encryption
arstechnica.com
arstechnica.com
Who are the Halliburton and Blackwater of the information security fields, or is it them all over again?
They are obviously superpac'g Clinton's campaign
No key? No data. Anything that can let government and industry collaborate to get in is an avenue that attackers can use to get in.
This is the sort of pointy-haired-boss thinking that gets parodied in Dilbert. Anything is possible, if you just throw enough money at it. Actual understanding of science and first principles is unnecessary if you understand that!
Instead, physicists hypothesized that atoms might undergo fission, which would lead to a release of energy and neutrons. Then they did further research into this matter and observed fission occur. Once this foundational research happened, not before, Openheimer drew a picture of a bomb.
It was only later, that the US spent about 0.2% of GDP on the research and development necessary to construct an atomic bomb.[0]
At this point in time, we have no hypotheses that are analogies to fission from which we can build the type of system that Clinton is suggesting. In fact, a solution to this problem would also be a solution to many of the problems that cryptographers have been researching for decades. Consider that under the proposed system, the Government has legitimate access to the encrypted data[1]. Additionally, we would like the system to prevent access to people without legitimate access. It will be a good day for crypto when we solve this weakened version of the problem at scale.
At the moment, the best solution to this problem we have is to decentralize all key material, so that any breach is localized. However, this solution is fundamentally contradictory to the proposed system.
Additionally, when we say that the Government is an honest actor, we really mean that the Government computer systems are honest actors. We know that this is not the case, because we have seen these systems be compromised by malicious actors.
[0] Total cost of the Manhatten project was about 2% of GDP but about ~90% went to construction, which would likely be far more expensive for a bomb than for crypto.
[1] With judicial approval, which I will ignore. For the purposes of this security definition, assume that the Government is an honest actor.
Anyone who wants the government not to see anything would just encrypt with the larger key as a standard (unless someone does a timing cheat in the decryption implementation itself).
There could be dirty tricks deployed, but truly breaking an encryption is a serious waste of time and money. It sounds to me that Clinton's tech team needs to get more than just policy wonks in there if we are to trust her with truly understanding technology.
Like the fact that H1 Visas are really a ploy to get cheap labor, and not that we are 'deficient' in tech know-how.
I think with quantum computing and possibilities that breaking encryption brings to the table goverments will have this in near future.
It's possible that quantum computing will offer a temporary way around this, but unless governments ban the sale of quantum computing, it's only a matter of time until quantum computing allows criminals the same ability. Also consider that if the government can use quantum computing to break encryption, what's to stop other governments like China from doing the same thing?