Le titre du Vif/de Morgen est un résumé un peu rapide…
Plus de détail, ici
Google teaches “AIs” to invent their own crypto and avoid eavesdropping | Ars Technica
▻http://arstechnica.com/information-technology/2016/10/google-ai-neural-network-cryptography
In conclusion, the researchers—Martín Abadi and David G. Andersen—said that neural networks can indeed learn to protect their communications, just by telling Alice to value secrecy above all else—and importantly, that secrecy can be obtained without prescribing a certain set of cryptographic algorithms.
avec lien vers le papier
►https://arxiv.org/pdf/1610.06918v1.pdf
LEARNING TO PROTECT COMMUNICATIONS WITH ADVERSARIAL NEURAL CRYPTOGRAPHY
Martin Abadi and David G. Andersen
-Google Brain
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.