Why quantum computers will pose a security problem
Marian Margraf, holder of Bundesdruckerei’s endowed professorship at Freie Universität Berlin, and his students are currently conducting research into quantum computer resistant encryption methods. He explains in our interview why this is an important topic for everyone, why time is of the essence and why quantum computers pose a threat to data security.
The search for quantum computer resistant encryption methods
Mr. Margraf, what’s so special about quantum computers?
Marian Margraf: In simple terms, classical computers calculate with zeros and ones whereas quantum computers use quantum mechanical states. They are physically entangled and can overlap. The advantage of this is that certain calculations can be performed much faster than with a classic PC. This means that large amounts of data, such as that generated in the age of ‘big data’, can be processed quickly. But there are also disadvantages: We now have a security problem.
In what way?
Marian Margraf: Interest in quantum computers has increased significantly in recent years. Many solvent interest groups, like Google or Alibaba, are conducting research and investing a lot of money. Experts estimate that this development will make quantum computers so powerful in about ten to fifteen years that they will be able to break the encryption methods used today.
This will have several consequences: This means problems for today's products – for instance, security hardware for VPN or security chips where conventional encryption methods are used. After all, today's processes will have to be used for a few more years before better alternatives become available. If these products are long-lived and if the predictions that quantum computers will break today's encryption processes in ten to fifteen years are correct, then security could become difficult.
In addition, the NSA already stores many data connections, which could then be decrypted retrospectively. If we want to make sure that the communications that we assume to be secure today will still be secure in ten to fifteen years' time, we do in fact need new encryption methods today. So, you could say: We need to hurry.
Why is it difficult to solve this problem?
Marian Margraf: Today’s cryptoalgorithms are broken down into symmetric and asymmetric methods. Quantum computers can weaken symmetrical procedures. This could be prevented by doubling the current key length. This is much more difficult with asymmetric procedures which cannot be saved simply be increasing the key length.
What will have to happen instead?
Marian Margraf: If we want to encrypt communications securely in the future and keep the content secret, we must look for new methods. Two aspects need to be considered here: For one thing, is the new method really resistant to the cryptoanalytical algorithms that can be implemented efficiently on quantum computers? Secondly, we have to find out whether the new procedure is also resistant to classical attacks.
What exactly are you working on?
Marian Margraf: We examine new cryptographic techniques that are proposed to us and we try to find attacks against them. A cryptographic procedure is considered to be practically secure if it resists all known attacks. This means we not only have to work hard, we also have to be very creative.
Professor Margraf, thank you very much for taking the time to talk to us.
Professor Dr. Marian Margraf
Marian Margraf is a Professor at the Institute of Computer Science at the Department of Mathematics and Computer Science at Freie Universität Berlin. He studied mathematics in Kiel. After completing his doctorate he first worked as a cryptologist in the Federal Office for Information Security (BS) and was responsible for introducing the German ID card from 2008 to 2012 at the Federal Ministry of the Interior. In 2014, he received the endowment professorship ‘ID Management Working Group’ of Bundesdruckerei at FU Berlin. He has also been head of the Secure Systems Engineering Dept. of Fraunhofer Institute AISEC since 2017.
Professor Dr. Marian Margraf, Professor at Freie Universität Berlin