Is post-quantum cryptography something we need to be looking at in the banking sector?

We hear a lot in the press about quantum computing, quantum cryptography and post-quantum cryptography (PQC), so what is the difference between them and how can they help the banking sector?  For this blog, I talked to our security specialists to better understand how this technology can make a real difference in my area; the banking industry.

What is post-quantum cryptography and why is it different from quantum cryptography?

Quantum Computing promises the ability to solve certain types of mathematical problems much more readily than classical computers can. Amongst these mathematical problems are the techniques we use to secure the vast amount of data stored in our digital world. This presents a problem for organisations that need to keep this data secure and that’s where Quantum and Post Quantum Cryptography come in. PQC is not really a branch of quantum computing. Rather it’s the development of classical computing cryptographic systems that are secure against quantum computing techniques as well as being secure against classical computing threats. Quantum cryptography is the use of quantum computing techniques to secure data and communications. This could include the use of techniques such as quantum key distribution, but that’s a topic for another blog!

Classical encryption techniques rely on the relative difficulty of solving certain mathematical problems, such as the factoring of very large numbers, thus making it practically impossible to decrypt data that has been encrypted using a sufficiently large key.

The challenge now is that quantum computing techniques can solve those problems much more readily than classical computers can. Thus, what was a problem that would take classical computers a huge, like the lifetime of the universe huge, time to solve can now be solved much more readily using quantum computing.

The reality though is that the currently available quantum computers are not quite large enough, or reliable enough to solve these problems for most types of encryption in use today, but the thinking is that they will be fairly soon, probably by the end of the decade according to Google’s research.

Are all classical encryption techniques vulnerable to quantum computing attacks?

One of the things that I have heard is that there is a difference in vulnerability between asynchronous techniques and synchronous ones.

In theory all classical encryption techniques are vulnerable to quantum computing attack, however, some are more vulnerable than others. Symmetric algorithms tend to scale well, in that they are highly performant and as computing power increases one can simply increase the size of the key makes the key derivation problem much harder to solve. It’s the very characteristics that make them suitable for say bulk encryption that also make them resistant to quantum threats. AES is a case in point.

Asymmetric cryptography, with a public/private key pair allows us to perform cryptographic functions between parties without sharing a secret. It’s this characteristic that underpins modern PKI services, blockchains and key exchange mechanisms. But the mathematics behind this characteristic and its lack of scaleability that makes asymmetric cryptography more susceptible to quantum computing attack.

Techy bits to consider for security

Most robust security models rely either on asymmetric encryption or a combination of symmetric and asymmetric encryption, using asymmetric encryption for the initial key exchange and then synchronous encryption using that key going forward. Obviously, this technique is vulnerable to the asynchronous key exchange segment being compromised, but the risk is reduced.

As a minimum, organisations should be looking to use larger keys for their synchronous encryption 256bit instead of 128bit for instance, but inevitably that’s a race to stay ahead of the current quantum computing infrastructure that means constant changes to your security tools.

On the plus side, quantum computers are really good at certain types of tasks, and not so much at others. For instance, they are really good at solving things like factoring problems, which means they can be used against a number of types of encryption today. However, they are not good at solving all types of problems, which provides an opportunity to develop cryptographic techniques that are not vulnerable to quantum computing techniques.

The National Institute for Science and Technology in the US has been working for the last few years on putting together a set of PQC methods that are not vulnerable to the sort of techniques that quantum computers can be used for. But these remain immature and some have been recently compromised.

Why should you care about this now?

Surely, we could audit our current cryptographic estate and implement sufficiently robust existing tools so that we don’t have to worry? That depends on the sensitivity and longevity of your data.

Let’s say that useable quantum computers, and the associated algorithms, that can crack most of the current encryption mechanisms are five to ten years away from being widely available. Depending on who you listen to that range is either very optimistic or very pessimistic, but the point is that it looks as though current encryption methods will be vulnerable to quantum computing attacks in the not too distant future.

So, the question organisations need to ask themselves is “is the data I have today still going to be relevant in a few years”? By that I mean would the data still be useful to a bad actor? Things like biometrics for instance won’t have changed very much from the data that you have already captured and stored, so if someone harvested that encrypted data today, then used some future application of quantum computing in five to ten years from now to decrypt the data, then it would still be useful to a criminal. The same could be true for financial records, medical records, sensitive corporate records and so on. Code signing and digital rights management being other examples.

The question that businesses, including banks, need to ask themselves right now is whether or not the data they have encrypted today is encrypted using appropriate techniques and whether it will still be relevant a few years from now.

What do we tell our banking clients?

A number of our clients have started looking at this problem, typically by auditing their current encryption estate to determine what tools and techniques they are using for encryption today and examining the sensitivity and longevity of the data across their estate. But many of them are asking is there anything else that they should be doing?

Well, as well as looking at the data and their current encryption techniques, we would recommend that they should also be working with their security suppliers, the people who provide the hardware and software to secure their data and look to see what they are doing to combat the challenges presented by quantum computing. Getting quantum ready!

Security system suppliers should by now have plans in place for how they are going to tackle post-quantum encryption and it would be concerning if they didn’t at least have a position and a roadmap as to how they plan to deal with this. This is a big area to understand and digest but we are here to discuss it all with you, so please get in touch. You can also find more on our quantum page.