There’s no doubt threat actors are either using or will soon use artificial intelligence to build malware and craft their attacks, says an expert.
“Bad guys almost certainly are going to use AI,” Dave Masson, Ottawa-based director of enterprise security at Darktrace told this week’s IT World Canada MapleSec Satellite webinar.
“Why? The same reasons we use AI: It’s going to do the heavy work for them. Likewise, if they can take the human being out of it [their work] and have the attack driven by AI, it’s less likely to make a mistake.
“In fact it’s going to be more efficient. It’s going to allow them to carry out more attacks at a greater scale than before.”
That means, he added, if adversaries are going to use AI to attack, CISOs have to find AI-driven solutions to defend themselves. “You can’t get into an AI fight,” Masson said, “You’re not going to win.”
Asked by session host Jim Love, ITWC’s CIO, if ChatGPT is a bigger threat than AI, Masson called it “a harbinger of things to come. It can write text they can use. If you’re an attacker where English isn’t your first language, you can kiss goodbye to all those silly grammar mistakes you make that allow your phishing emails to be intercepted.”
On the other hand, Masson added, it may be possible for an AI solution to detect if the text of an email message was been generated by a human or not.
Related content: Does AI in IT operations pay off?
Artificial intelligence solutions can be used in many scenarios, he said, from discovering assets on IT networks to reacting faster than humans to quickly spreading malware.
An AI solution may be able “heal” your environment by getting it back to the condition it was in before an incident, Masson added.
Not that AI is perfect, he stressed. “It’s got to work on the data it has,” Masson said, and that data is input by humans. As the old saying goes, he said, “it’s garbage in, garbage out.” The best way AI can be used is not by trusting its responses completely, but by using it to help IT/infosec staff make decisions, he said.
“Some people say, ‘It’s AI. Should I care?’ You should care. What you shouldn’t do is worry, because this is pretty much the way we’re going. It’s all about using artificial intelligence in all its different modes and guises to support human beings.”
The next MapleSEC Satellite online series event will be held June 21.
