Podcast Transcription


Christine Izuakor:
Welcome to the Veriato Insider, a biweekly podcast covering some of the latest trends and things to know in cybersecurity. This podcast is sponsored by Veriato, a next generation employee monitoring and insider threat detection software provider. To learn more about how Veriato can help protect your company, check out Veriato.com. I'm Dr. Christine Izuakor, your host for today's segment, and we're talking about an area of insider threats that I don't think gets enough attention, and that is how artificial intelligence and user behavior analytics can support physical safety, especially in the work environments, through advanced predictions on dangerous and even potentially life threatening situations.

Christine Izuakor:
Now oftentimes in the industry, people get so caught up, you know, focusing on the basics of confidentiality, integrity, and availability. Which are super important, right, but the downstream impacts, when you look beyond protecting data, are also significant. And so today's topic is all about how advancements in technology can help entities, again, predict and detect physical insider threats. Uh, joining in on this conversation we have a remarkable guest with us today to cover this. Someone I've looked up to my entire life. Her name is Teka Izuakor. Uh, now Izuakor isn't a common name (laughs) and so if you are already wondering, no, that's not a coincidence. Uh, we are sisters. And also proud fashionable tech geeks.

Christine Izuakor:
Um, (laughs) I've been holding down the fort more on the cybersecurity side while she's focused a lot on the artificial intelligence and data science side. And so the perfect dynamic duo for this conversation. Um, a little bit more about Teka, she's an AI and data science researcher with experience unraveling how rapid advancements in machine learning impact areas like critical infrastructure. Um, she spent about 10 years working in a variety of business domains over the last period of time within large organizations like Shell Corporation, Stolt-Nielsen, and more. After earning her MBA from Texas A&M University, she began focusing on machine learning projects, and is currently wrapping up a professional certificate in artificial intelligence at Columbia University.

Christine Izuakor:
Welcome, Teka, and thanks for joining us.

Chika Izuakor:
Thank you so much for having me. I'm super excited to have this discussion today. And, uh, aside from the fact that you're my sister, I definitely look up to you as well, in everything that you've done to create awareness in the cyber security space. So thank you so much for inviting me.

Christine Izuakor:
All right, it's my pleasure. We're so happy to have you. So to start out, tell the insiders more about who you are, your background, and your journey into AI.

Chika Izuakor:
Sure. So I don't really have the traditional path, um, into tech that most people who are considered techies do. I actually started in fashion and then transitioned to oil and gas. Um, in the business space. And then through that, actually started trading options and became obsessed with the stock market. And funny enough, my interest in AI came from just personal, uh, curiosity on how I could become a better trader. And so in my, in my studies on that and, and algorithms, and how machine learning is doing amazing things in the trading space, I started trying to apply it to other areas of my role, um, that I had within other organizations. And, and just being able to see how AI is impacting and will impact every industry, um, has just heightened my level of curiosity and my interest in the field.

Chika Izuakor:
And so once I kind of decided that AI would be helpful as a trader and within the other business areas that I was involved in, I started on my own really just seeking out more knowledge, and then transitioning to completing the professional certification and just really this never ending search for more information and, and just learning the ways that we can apply AI in our every day life, and to solve some of, um, our, our toughest challenges. So anyways, here we are.

Christine Izuakor:
High demand in super important area of artificial intelligence and machine, uh, machine learning, which is so important. And so jumping into our topic a little bit more, we know insider threats come from a variety of forms and, and fashions from unsuspecting employees who introduce a risk to companies to criminal insiders who steal from companies and are malicious in that nature. And so I know a ton of ways that AI, um, is helping to shift this space, and would love your perspective. But first, in layman's terms, maybe help us all understand the significance of machine learning and even user behavior analytics, especially when it comes to insider threats.

Chika Izuakor:
Yeah, so user behavior analytics allows any organization to collect all of this information over thousands of employees and package it in a way that helps them to prevent threats. So for example, if you have an employee who signs in and does a particular thing every day, others may not be able to see if there's small changes, um, like in that person's behavior online. But AI can help to detect that maybe, this person's signed in, but they're not doing what they would normally do. And so that could be an indicator of a breach or threat. You have all these little threats that are coming from everywhere. And so there are great cybersecurity tools that help to detect those.

Chika Izuakor:
But when you're trying to look at 40,000 people and all of their behavior, and how all of that could, um, could compromise your systems. Or how people who have malicious intent can break into your system or cause data breaches, um, it's difficult to apply human resources to that amount of people and to catch all of the little nuances in everyone's personality that may be clues. So machine learning and AI collects all those clues and spits out this information that says, okay, based on what this person's doing right now, or maybe what this person has done over the last three weeks, there may be a threat. And so being able to predict what people are going to do before they do it is, is just huge in preventing cyber attacks.

Christine Izuakor:
So I completely agree on the predictive side. Especially, I've heard a lot of people relate it back to, um, you know, finding a needle in a haystack that of course a human being would, wouldn't be able to, uh, find. But having all of the data and all of the, uh, insight over time, and the context around activities, can, uh, make that difference in finding things that of course people wouldn't. Now, we've seen interesting, uh, use cases around user behavior analytics and safety surface over time as well. And, in some cases, insider threat, we know, can be life or death. Which, like I mentioned in the beginning, is something that a lot of people don't focus on.

Christine Izuakor:
Um, a lot of people focus on, uh, the data side of things, from a confidentiality, integrity, availability standpoint. But you know, for example, this is, uh, a very grim reality. According to the FBI, and one report that I saw in the last two decades, there were nearly 1,000 people who were, unfortunately, killed, and 1,500 people injured during, um, active shooter incidents in the US. Um, and last year, about 20% of those occurred in commercial spaces. And so I think that's just a, a testament to how detecting insider threat is important, not just from a digital standpoint. But the physical threats matter as well, right?

Christine Izuakor:
And so, my two questions for you are first, what role can, uh, user behavior analytics and artificial intelligence play in detecting threats like active shooters, or let's say suicidal individuals, for example? And the second part is, what are some specific behavioral indicators of compromise that, uh, machine learning can pick up on that human beings or traditional technology can't?

Chika Izuakor:
Yeah, those are great questions, and those numbers are, are just so scary. But you know, when, when people think about AI and, and safety, and active shooters, and, and anything that has to do with law enforcement, they think, like, RoboCop. But the reality of it is, there are other tools that are, that are so beneficial when it comes to stopping active shooters. Like image detection. Um, image detection and AI works in, in such a great where, where, through outside cameras, they can scan for objects such as weapons, and also scan people's faces against databases that show people who've committed previous crimes that are weapons related. So things like that are really helpful in the safety space when it comes to threat detection.

Chika Izuakor:
I've also seen some really amazing things that allow organizations to pick up on certain keywords in emails, um, that may give an indication that an employee is disgruntled. Um, I think that when we're in a work environment, you know, we talk, we're around a water cooler, we may say different things. But what you don't tell your employees, you tell your computer sometimes, right? So based on your Google searches, the things that you like on social media, your posts. Um, all those things that you do on, online to create this digital footprint. And sometimes, that can create a pattern that indicates potential, um, behaviors that, that may be detrimental to the organization or the employees there.

Chika Izuakor:
Um, they may even be indicators of potential suicide, because you know, so many teens won't talk to their parents, won't talk to school officials. But they'll have conversations online, and those conversations, through AI, can give us clues as to who's more at risk. And so AI does some really great things in that space. Because again, like you said, the needle in a haystack, there's this, there's this, uh, human ... There's the human element, for sure, where you, you notice that someone is maybe acting differently or erratically, right? But what happens when you don't notice that? A machine can notice what a human can't, right?

Chika Izuakor:
And so, whatever they don't tell you, you figure out from those analytics that you get from their behavior. So those are really key. And I mean, it, it goes past emails and keywords, and, and delves even deeper into the tone of certain words can provide a, a clear indicator that there's an employee who's maybe having a repetitive issue with a certain person. And so, those may be precursors to more violent actions. And if we can identify those things with AI, um, it, it definitely makes the playing field more level in that we're, we're not so focused on using human resources to detect those things.

Chika Izuakor:
Again, we're being proactive by having this machine learning take all of this data across thousands of employees, which would be so cumbersome for humans, and just almost impossible to really attempt. Um, but to, to scan all this de- data and package it in a way that says, okay, this person may have some suicidal tendencies or thoughts based on this behavior that we're seeing online, um, that's just huge. And I, I think it's so important.

Christine Izuakor:
Yeah, it's definitely huge. And for some, probably overwhelming. And so as we wrap up here, um, do you think that companies are ready to embrace this type of technology? And, and are there any final thoughts you wanna share as well?

Chika Izuakor:
Oh man. So, uh, you know, when people think about AI, when companies typically think about AI, they think about the flashy stuff that you see at the expos, you know? But the truth is that most companies are just so starved for the fundamentals around data and conditioned learning that the thought of threat detection through AI just seems so far away, because they're in the infancy stages just trying to build these teams, and, and learning to use data and AI in their day to day operations to help create revenue. It's definitely on the horizon, I think, as, as we bring more awareness to the benefits of, of using and implementing AI, companies will get ready. You know, because it, it's coming.

Chika Izuakor:
And some companies, for sure are, um, way more advanced than others. But, but I think most companies are just in that space where they're just trying to build the ground work, and even, you know, figure out what exactly they need. So it'll, it'll be interesting to see what the future holds, as far as implementing AI into the day to day operations, and, and how they are going to be using it for threat detection, if they get to that point.

Christine Izuakor:
Yeah, I think the awareness piece is so key. Because to your point, like it's one thing to be, you know, super advanced, and already know what data you have in, in house and how that might see different algorithms. And then all of that. But it's another thing to barely even have a, a handle on your data, because you don't know where to begin. I think that one of the interesting things for the cybersecurity industry would be ... There's a, a lot of good solutions that are popping up and vendors who are really building expertise and a lot of capabilities anchored around AI that I think companies can benefit from without having to do a lot of that grunt work on their own.

Christine Izuakor:
And so I think the natural progression that we'll see is, uh, companies begin to embrace it through, you know, third party solutions, and tools from, you know, vendors who are already investing a lot in that space in the cybersecurity industry. And then, we might see that continue to progress to people starting to bring things in house, and figure out how they can get more comfortable with, um, integrating some of these things into their, you know, development life cycles and all these kinds of things. So, um, really good, really good insight there.

Christine Izuakor:
So, um, Teka, thanks so much for joining us today. That concludes the Veriato Insider podcast for this week. Again, this podcast is brought to you by Veriato, an award winning cyber security company recently recognized with the Gold Award for Best Insider Threat Solution of 2020. Their solutions are anchored around four core pillars of cyber security protection, including employee monitoring and web filtering, the, workplace investigations, insider threat detection, and ransomware support. So to learn more about how Veriato can help protect your company, check out Veriato.com.

Christine Izuakor:
Thanks for tuning in, and a special thank you to Teka, again, for joining us today.

Chika Izuakor:
Yeah, thank you for having me.

Christine Izuakor:
I'm Dr. Christine Izuakor, the CEO of Cyber Pop-up, and it's been our pleasure to share these insights with you. So until next time, stay safe and secure, insiders.


Listen To Podcast