Skip to main content

Palo Alto director: "We could see AI in antivirus in the next 5-10 years"

Home automation, connected objects, e-health and quantified self, bio-informatics and Web 4.0... All these are part of the emerging trends of the future, and innovation is constantly progressing.

But the security of these new technologies is not always anticipated.

At the International Forum for Cybersecurity (FIC), an elite panel formed of industry-leading experts from the top cyber security companies in the world seeks to answer the question: How can we consider the security and uses of the future?

Isabelle Dumont, the director of Industry Solutions at Palo Alto Networks shared her predictions, fears and hopes for the future of security.

What are the dangers of the growing Internet of things?

The benefits overall of those connected objects are so much superior to the perceived risks of big data and connected cities. We shouldn't fight that, but there needs to be an awareness spreading among the population, Hopefully we won't need to wait for a cyber war or a major scandal story to come out before that awareness spreads

So people basically don't bother with security?

Yes, people are often not changing their passwords frequently; firewalls are not updated, and people are sometimes not putting them into place in the first place due to their cost and the effort required to set them up.

That's why it's the role of the government to spread the news, and get people's awareness up to an acceptable standard.

What do you foresee in the future of advanced threat detection?

We're looking for the proliferation of certain tools, and the growing complexity of many processes. In 5-10 years we imagine most security being handed over to automated processes. That means being able to spot abnormal behaviour and abnormal processes on an automatic basis, and only involve human input in a small percentage of attacks.

It's about setting up a specific architecture capable of preventing attacks by emerging threats. We're looking at utilising artificial intelligence and machine learning more and more, but that won't be for a couple of decades maybe. We need greater advances in AI before that can truly be useful. It will be a slow transition.

Surely there's a human element to secure as well, though?

It's true that most hackers nowadays aren't machines or automated systems, but humans. We need to increase the rate at which the private sector adopts security and fights against cyber crime.

It's also true that there are problems with the collaboration between security companies and other enterprises.

There are consortiums being put in place to allow companies to share security information without putting that information necessarily in the public domain.

More from FIC 2014:

Prominent lawyer: Does the cloud have international borders?

US Navy professor defends PRISM, calls Internet "a lawless frontier"

French defence minister calls for unified front against cyber crime

Estonian IT security chief: "I don't want to use American encryption anymore"