Adversarial algorithms to be a key theme at CyberSec & AI Prague conference
CyberSec & AI Prague will bring together global experts from the fields of both artificial intelligence and cybersecurity this October.
One of the major attractions for those attending the event is that a significant amount of the conference content addresses the intersection at which the two disciplines of AI and cybersecurity converge.
As addressed in a recent Avast blog by Rajarshi Gupta (Head of AI at Avast) and Sadia Afroz (AI Researcher at Avast and International Computer Science Institute, Berkeley), a key topic of CyberSec & AI is adversarial AI.
“Defenders train our AI classifiers to be more robust by feeding them many examples of good and malicious files,” write Rajarshi and Sadia. “This active arms race makes AI in security particularly challenging. We must teach our AI to look for those disguised threats, and always stay a step ahead.
“One way to do that is by generating adversarial examples to teach our classifiers by using approaches like generative adversarial networks (GANs). Another way would be to model ‘good’ or ‘acceptable’ characteristics, which might be possible to protect simple devices like IoTs. The problem is the adversary has access to the same AI as the defenders.”
Sadia and Rajarshi argue that AI still relies on security through “obscurity”, meaning the only way to protect AI is by “hiding it from the adversary”. You can read the full blog here.
For those keen to learn more on this fascinating topic, Sadia will be presenting her talk ‘Recent Advances in Adversarial AI for Malware’ at CyberSec & AI on the 25th October at 12:30.
Rajarshi will also be presenting at the conference, his presentation is entitled: ‘Security is a Unique Challenge for AI, and Deserves Stronger Research Focus’.
To be part of CyberSec & AI Prague 2019, secure your tickets here.