AI to Detect and Defend Against Cyber Attacks

Watch the 4-minute video here or on Youtube, or read the transcript below.

Transcript

dotmagazine: How is artificial intelligence being used to help to detect cyber attacks and to defend against them?

Marc Wilczek: Well, in the digital age, time has become the real critical issue here. And it’s all about ensuring instant responses, because we clearly see human error as one of the biggest issues. It just takes humans too long to detect cyber attacks and to respond to it. And that’s clearly an area where AI can add value, fundamental value, by being able, for instance, to determine what is your regular normal traffic that is going to be translated into a data set that says "this is your typical data set". And once there is, let’s say, a malicious attack or something awkward happening, clearly AI can augment human interaction by clearly understanding and detecting instantly that this is something suspicious, and perhaps block that, in order to prevent an issue rather than allowing humans to respond to an issue. So, it’s kind of a paradigm shift. It’s the attempt very much to prevent an incident from happening as opposed to responding to a cyber attack.  

Wilczek: Very good question. I think it is absolutely essential, to put it that way. And it is essential for a number of different reasons. First and foremost, there is this mega talent shortage in the marketplace. Some studies predict that there is going to be 3.5 million vacancies for cyber expertise in two years down the road. So the issue is that demand outweighs supply very fundamentally, so people are just not there. And that’s one of the issues why organizations and users need to rethink their defense strategy, and clearly AI is an area that can address that talent shortage to begin with.

Beyond the talent shortage aspect, as I’ve just said, the timing issue is another area that clearly underlines the necessity for leveraging AI, simply because when being used in combination with automation and machine learning, AI can just be instantly fast. It can be so fast, far beyond a human. And this is again an area why and how AI can add value. 

dot: What legal or technical implications need to be considered when setting up an AI-supported security system? 

Wilczek: So, I think the first thing to begin with is that, when rolling out that AI-based system, the learning period initially is the most critical path. So what you need to do is to really make sure that you’re feeding your AI-based system with clean and legitimate data. The equation is relatively simple – it’s kind of garbage in, garbage out. Or, in other words, if you put good data in – clean and legitimate data – you get the right type of responses. Otherwise, if the system is being fed with wrong data sets, the system will ultimately fail and make wrong decisions. So you want to make sure that during that learning period, there is only clean and healthy input being put into the system. 

Marc Wilczek is a business leader with more than 20 years of experience in helping market-leading ICT companies to transform themselves, expand into new fields of businesses and geographies, and accelerate their growth. He currently serves as Managing Director at Link11, a leading IT security provider that ensures cyber resilience for its customers with a 360 degree, AI-based cloud security platform.