Learning to Protect Communications with Adversarial Neural Cryptography

By Martín Abadi, David G. Andersen. 2016

235 citations recorded by the source corpus.

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.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Safety, Robustness & Interpretability, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Safety, Robustness & Interpretability

In Signal Processing

Citations in the corpus

Generalization and Equilibrium in Generative Adversarial Nets (GANs) (2017, methodology): “Generative Adversarial Networks (GANs) [Goodfellow et al., 2014] have become one of the dominant methods for ﬁtting generative models to complicated real-life data, and even found unusual uses such as designing good cryptographic primitives [Abadi and Andersen, 2016].”
NIPS 2016 Tutorial: Generative Adversarial Networks (2016, neutral): “• Adversarial mechanisms for cryptography (Abadi and Andersen, 2016).”
Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments (2017, methodology): “This is similar to the cryptography environment considered in [2].”
An Introduction to Deep Learning for the Physical Layer (2017, background): “However, in [50], the concept of learning information encoder and decoder which adapt to a variable key to provide privacy is introduced.”
Deep Learning Based Communication Over the Air (2017, background): “ing of encryption/decryption schemes for an eavesdropper channel [24], as well as compression [25].”
HiDDeN: Hiding Data With Deep Networks (2018, background): “Abadi and Anderson [23] show that adversarial networks can be trained for cryptography; Uchida et al. [24] embed watermarks into the weights of trained neural networks; Fang et al. [25] use LSTMs to embed secret messages into text.”
Generating steganographic images via adversarial training (2017, methodology): “Abadi and Andersen [2] used adversarial training to teach two neural networks to encrypt a short message, that fools a discriminator.”
Learning Generative Adversarial RePresentations (GAP) under Fairness and Censoring Constraints (2019, methodology): “Adversarial learning approaches for context-aware censoring and fairness have been studied extensively Edwards and Storkey [2015], Abadi and Andersen [2016], Raval et al. [2017], Huang et al. [2017], Tripathy et al. [2017], Beutel et al. [2017], Madras et al. [2018], Zhang et al.”
Deep Learning in Mobile and Wireless Networking: A Survey (2018, background): “Recent research reveals that GANs can even protect communications by crafting custom cryptography to avoid eavesdropping [464].”
Stabilizing Adversarial Nets With Prediction Methods (2017, unknown): “…Wang & Gupta, 2016; Isola et al., 2017), domain adaptation (Taigman et al., 2017; Tzeng et al., 2017; Ganin & Lempitsky, 2015), imitation learning (Ho et al., 2016), privacy (Edwards & Storkey, 2016; Abadi & Andersen, 2016), fair representation (Mathieu et al., 2016; Edwards & Storkey, 2016), etc.”
Generative Adversarial Privacy: A Data-Driven Approach to Information-Theoretic Privacy (2018, methodology): “Both the privatizer and the adversary in the GAP framework are trained on Tensorflow (Abadi et al., 2016) using Adam optimizer with a learning rate of 0.”
Context-Aware Generative Adversarial Privacy (2017, background): “Our work is also closely related to adversarial neural cryptography [84] and learning censored representations [85], in which adversarial learning is used to learn how to protect communications by encryption or hide/remove sensitive information.”
A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications (2020, methodology): “GANs have been widely used in many other areas such as malware detection [434], chess game playing [435], steganography [436]–[439], privacy-preserving [440]–[442], social robot [443], and network pruning [444], [445].”