A third of experts predict cybersecurity won’t need human decision-makers within a decade: This Week in AI

A third of experts predict cybersecurity won’t need human decision-makers within a decade

According to a research paper published this week, experts studying the industry believe the rise in Artificial Intelligence that we’ve been seeing lately will only continue to evolve and grow until there is no need for humans to be involved in the cybersecurity decision making process in less than ten years.

Another 13% said that they believed it would take longer than ten years and 14 % said there was already no need for humans to be making decisions due to the rise of AI in cybersecurity.

The remaining 40% believe AI will never become advanced enough to eliminate humans from making decisions in regards to cybersecurity.

Right now, cybersecurity professionals are in short supply despite the ever expanding growth of the field. The use of AI has seed rapid proliferation recently to try and automate the most boring and repetitive cybersecurity processes and lessen the amount of work for security teams.

Machine learning has been particularly successful at identifying when hackers are attempting to breach a network, in ways a human being could not. Read More

Knowing Normal with AI in Cybersecurity

The way that most effective AI works is by creating a baseline of your network and in doing so putting together an idea of what ”normal” looks like for your network as opposed to what might be a threat that’s messing up regular traffic patterns on your network.

This week Sydney J Freedberg Jr. wrote a piece for Networks & Cyber articulating the common flaw of most security systems.

“How do I even know what’s normal and what’s abnormal so I can detect anomalies? We simply don’t know,” says Dean Souleles, chief technology advisor for the Director of National Intelligence.

Luckily there are emerging companies like MixMode that are using unsupervised learning Artificial Intelligence which is capable of precisely knowing which network operations are normal and which are worth flagging as threats.

With supervised learning, the problems presented in this article continue to exist because the intelligence can only detect what it has learned to be threatening through past experience and has labeled as such, whereas with Unsupervised Learning based AI, the system is able to take in the context of a situation and understand whether or not something is a threat based on the circumstances. Read More

Artificial Intelligence To Play a Key Role in Defence Cybersecurity

Air Force Lt. Gen. Jack Shanahan, director, Defense Department’s Joint Artificial Intelligence Center said that Artificial Intelligence will play a key role in assisting cyber operators to make speedier and more accurate decisions while participating in a panel at the 10th Annual Billington Cybersecurity Summit in Washington yesterday.

Apparently the DOD is becoming increasingly focused on AI — and the possible capabilities of counter, or adversarial AI — the general said. However this will only work if those tools are able to function in extreme environments and are trusted by those who employ them, he added.

“Counter-AI is in the future. It’s analogous to electronic warfare,” the general said. “That’s something that’s upon us now, and we need more thought put into that commercial enterprise.”

When asked which challenges he has come across he mentioned the difficulty of managing and monitoring big data.

“What does normal look like so anomalies and variances in the system can be detected in the data?” he said. Read More