Understanding the role AI might play in cybersecurity

AI is like the human mind — it'll need to be trained quite extensively before it can start making a real difference in the industry
19 October 2018

Can AI really make cybersecurity better?. Source: Shutterstock

Today, cybersecurity is a priority for every business — however, despite their best efforts, defending the company from threats is becoming harder.

The industry faces a talent shortage, has more end-points to defend as the internet of things becomes a reality, and at the same time, is seeing attackers get smarter and leverage new technologies.

As a result, experts believe that artificial intelligence (AI) will play a big role in cybersecurity. Not immediately, but definitely.

In the near future, when machines take up the tasks that need consistency, and people can be given the time to bring creativity and imagination to the table — that’s when we’ll have the ideal partnership between human talent and AI.

In reality, although having the constant, dependable support of machines sounds like a great idea, achieving it can be a tough task as cybersecurity isn’t as manual or repetitive as many of the other fields where AI has already made a significant contribution.

The challenge of developing cybersecurity AI

Scientifically or theoretically speaking, an AI algorithm is just like a human mind. It needs to be trained before it can be tasked with something as serious as defending the company against external threats.

Now, training is of two types. Training based on historical data and on the job training.

Using historical data to train an AI algorithm for cybersecurity is important, but it’s not going to help achieve much as attackers employ new and innovative ways to gain access to company assets and vulnerable devices.

The only other possibility is what people ordinarily call ‘on the job training’ — and it’s what we’ll need in order to get started with training AI for cybersecurity. Unfortunately, this is an endeavor that will take several man-hours before anything tangible is achieved, but it’s an important end-goal.

Ultimately, the idea is to train AI to help spot anomalies, irregularities, and vulnerabilities quickly and efficiently, irrespective of the creative methods employed by attackers.

Experts believe that in the near future, AI will be able to alert cybersecurity professionals to these issues before they can be used to cause damage, strengthening the company’s cyberdefense before and protecting it from attacks — day or night.