Fully homomorphic encryption, or how to perform operations over encrypted data | Orange Research blog | cross pond high tech | Scoop.it

Can we outsource medical analysis without giving away our medical information? Can we do biometrical identification without revealing our characteristics? Can we make statistics on data that we do not know? Yes we can, thanks to a cryptographic mechanism called “homomorphic encryption”.

Cryptography has known many transformations over the years. Many centuries ago, it was first used to protect military and political communications. Though very simple, the mechanisms then devised are still the foundation of current cryptography. The introduction of the computer during Second World War considerably increased the computation capacity. This increase reflected on cryptography in the late 70’s, when public key cryptography was invented. Cryptography became a thriving scientific field. Numerous academic works were produced, commercial standards were set and cryptographic algorithms began to secure our daily life. Today, cryptography is everywhere: in our credit cards, in our phone communications, in our internet browsing, etc.
But new services are today under deployment, such as mobile services, cloud computing, BigData or IoT. These services generate and process a huge amount of personal and sensitive information. As users become more and more concerned about their privacy, and industries want to protect their sensitive data, a new challenge arises for cryptography. Indeed, if this data was to be simply encrypted, processing it would be impossible. This leaves users and service providers with a dilemma: choose between usability and confidentiality of these sensitive data. Here comes fully homomorphic encryption!