Episode Summary: When Google’s DeepMind won against one of the best modern Go champions, is used multiple AI approaches and exposed gaps in some individual strategies. This even has shed more light on AI, but also on the utility in combining approaches to AI for individual problems. Data security is one of these problem areas where multiple AI approaches is being used to make our information safer. Dr. Sal Stolfo has been a professor at Columbia in Computer Science since 1972 and is now also the CEO of Allure Security, with a focus on engineering network intrusion detection solutions using AI applications. In this episode, Stolfo talks about the various styles of AI and statical methods that have been and are being used to detect malicious activity, as well as how he believes the future of security is going to have to adapt as increasing amounts of data become available.
Expertise: Parallel computing and database inference and computer security
Brief Recognition: Sal Stolfo is professor of Computer Science at Columbia University. He received his Ph.D. from NYU Courant Institute in 1979 and has been on the faculty of Columbia since. He has published several books and well over 250 scientific papers since, and received several winning best paper awards, in the areas of parallel computing, AI knowledge-based systems, data mining and most recently computer security and intrusion detection systems. He has been granted 33 patents in the areas of parallel computing and database inference and computer security; most have been licensed or sold. His research has been supported by DARPA, NSF, NSA, CIA, and several other companies and state agencies. His most recent research is devoted to payload anomaly detection for zero-day exploits, secure private querying, private and anonymous network trace synthesis for Predict.org, symbiotic embedded machines, automatic bait generation for trap-based defense to mitigate the insider threat and multi-core parallel computing.
Current Affiliations: Professor of Computer Science at Columbia University
(2:11) Being someone who started off in the research space – what types of, or applications of, AI are really most prevalent in security today?
(5:41) This AI that we’re talking about (deception security), it sounds like if we’re just duplicating a file it’s not, but it sounds like you’re (rather) creating a system…
(8:48) So when you have Chrome open…when it is you, what programs are generally running, and what do you do when your mouse when something’s loading – is it all of these various micro-activities about how one generally works (that is active authentication)?
(9:45) I imagine determining the wheat from the chaff on that is an interesting process, i.e. which particular facets of program use and mouse gestures…are really indicative of an individual, and which have too much overlap…this was probably a pretty big sift for you…
(12:58) Allowing the machine to pick up on what are the most genuinely differentiating factors from person to person and allowing it to really snip out those meaningful patterns…would I be correct in saying this is an unsupervised activity in some regard?
(18:38) We’re talking about AI approaches leveraged successfully or unsuccessfully for the purposes of security…what sorts of malicious AI exist out there in the world, what kinds of approaches are becoming more popular for “the bad guys”?
(21:35) Where does game theory play into AI and how do you see the crossover?
(24:16) It sounds like you’re of the belief that the thought of my chess move, your chess move…shouldn’t be an accepted norm, but that we ought find ways to…squelch ongoing attacks in a more effective manner…
Related AI and CyberSecurity Interviews/Articles:
- Artificial Intelligence and Security: Current Applications and Tomorrow’s Potentials
- Machine Learning Cyber Security May Help Speed Response to Hack Attacks (Igor Baikalov, Chief Scientist, Securonix)
- Artificial Intelligence’s Double-Edged Role in Cyber Security (Roman Yampolskiy, Professor at the Speed School of Engineering, University of Louisville)