Learning to Protect Communications with Adversarial Neural Cryptography

Abstract:

We ask whether neural networks can learn to use secret keys to protect
information from other neural networks. Specifically, we focus on
ensuring confidentiality properties in a multiagent system, and we
specify those properties in terms of an adversary. Thus, a
system may consist of neural networks named Alice and Bob, and we aim
to limit what a third neural network named Eve learns from
eavesdropping on the communication between Alice and Bob.
We do not prescribe specific cryptographic algorithms to these neural networks;
instead, we train end-to-end, adversarially.
We demonstrate that the neural networks can learn
how to perform forms of encryption and decryption, and also
how to apply these operations selectively in order to meet
confidentiality goals.

Conflicts: google.com;cmu.edu

TL;DR: Adversarial training of neural networks to learn rudimentary forms of encryption with no pre-specified algorithms

Community Implementations: [ 5 code implementations](https://www.catalyzex.com/paper/arxiv:1610.06918/code)