November 1, 2016

AI is the next security battlefield

Markus Melin

Head of Tieto Security Services, Tieto

Barack Obama, Bill Gates, and Stephen Hawking are some high profile public figures who have expressed their concern about Artificial Intelligence taking the place of humans in decision making.

Others have a more positive approach and say that AI is just an extension of human thinking that can make intelligent decisions faster than us regular people.

Just some days ago, Tieto nominated an AI member on the leadership team of a business unit, called Alicia T. And most of us are already familiar with personal AI assistants such as Siri, Cortana, Alexa, and Google Now.

We’ve truly come a long way since the pesky but strangely charming Office assistant Clippy offered us “help” when writing a document in Microsoft Word.

Whichever way you look at it, AI will be an extremely important part of all IT domains, and this means also cybersecurity.

Cat and mouse continues

IT security has often been described as a game of cat and mouse. The bad guys try to devise attacks that cannot yet be detected, while the good guys try to figure out beforehand where the next attack could be coming from and hitting to.

This is also a playing field where AI is going to play an important role. How will intelligent bots change security? Are they more of a threat or a remedy? And maybe most importantly, will the cat and mouse chase carry on as before?

Right now, it looks like the good guys are winning. MIT has recently built a system which can allegedly detect 85% of cyberattacks. And as with all intelligent machine learning systems, it will also get better over time as more data is fed into the system.

At this year’s DEF CON conference, there was a challenge where automated security systems competed against each other to see which could detect and patch software vulnerabilities most effectively. The winner was awarded a prize of 2 million dollars. It pays to be good at security!

The war of the robots?

Of course, also the attackers have AI tools and methods at their disposal. The same way defenders can use AI and algorithms to predict and prevent attacks, also attackers can gain from using clever automated approaches to circumvent security.

Will we be faced with an onslaught of bots with criminal intelligence, then? For Example, Patrick Hubbard from IT management software company SolarWinds says that the AI-based attacks are lucrative because they offer far greater sophistication, scale, and ROI than traditional cyberattacks.

It’s actually quite thought-provoking to imagine a situation where intelligent bots on both sides of the fence learn from each other’s actions and wage a never-ending war. However, manyexperts (myself included) believe that AI will be able to protect us better from attacks and turn into a force for good.

Many security companies with brilliant researchers are creating fascinating defence strategies. For example, British Darktrace is modelling their approach after the human immune system.

In cybersecurity, AI will be your friend.

Read more about our security approach here.

Stay up-to-date

Get all the latest blogs sent you now!