How advanced analytics protect Azure from cyberattacks

Azure Government CISO Matthew Rathbun and Relativity CSO Amanda Fennell explain how machine learning adapts to cyberattacks and protects the cloud.

TechRepublic spoke with Azure Government CISO Matthew Rathbun and Relativity CSO Amanda Fennell to provide insights on the role of artificial intelligence (AI), machine learning, and deep learning play in amplifying the company's cyberdefense posture. Below is a transcript of the interview.

Rathbun: Yeah. Absolutely. So, from our perspective we're collecting those 7 trillion events every day. We're running those through typical signature based analysis that you'd expect as part of the standard response protocol. We then also put all of that rich data into a cold bath model where we have advanced machine analytics, deep learning, neural nets, and that's looking for patterns that we didn't already see, that we weren't already aware of.

Fennell: And I guess it goes to say that obviously we have an eDiscovery solution that's capable of doing machine learning and AI on its own, that obviously our security approach is going to be similar. So, we also are trying to take as much as we can to automate it. It feels like those are a lot of the buzzwords in the industry when you say like, "Automation, machine learning." But the reality is we're just trying to get to the bottom of the line, like he said, where we automate and learn from the things that are coming at us so that we don't have to have a person sitting behind it the next time. So we try to focus on that as well. He mentioned heuristics. Obviously that's a big part of it for us, but the signature development that we do is where we try to respond accordingly. Probably, that's the area that we have the most fun in, is trying to get this responsiveness to the threats that are coming at us and developing our response to it so that the next time a human's not behind it.

Also see

Image: iStock/Yozayo

About Dan Patterson

Dan is a Senior Producer for CNET and CBS News.

Editor's Picks

Free Newsletters, In your Inbox