Where AI meets cybersecurity: Opportunities, challenges and risks so far

As evidenced by the instance of ChatGPT, synthetic intelligence is advancing in unprecedented instructions to resolve thrilling new issues.

However, as AI is being pointed towards vital cybersecurity operations, do the positive factors outweigh the potential dangers and issues?

“Completely, you need to be frightened,” stated Andy Thurai (pictured), vice chairman and principal analyst at Constellation Analysis Inc. “The issue folks don’t notice is that ChatGPT, being a brand new, shiny object, it’s all of the craze that’s about. But the issue is that many of the content material that’s produced both by ChatGPT or others are property with no warranties, accountability or in any respect. If it’s content material, it’s OK. However whether it is one thing like code that you just use, then it’s principally not.”

Thurai spoke with theCUBE business analyst Dave Vellante at CloudNativeSecurityCon, throughout an unique broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They mentioned the potential moral issues of mixing AI and cybersecurity and their ramifications.

Digging deeper than the floor

Whereas many within the software program and computing industries are overjoyed on the ramifications of AI for builders and engineers that imply properly, there’s one other, extra sinister aspect: The truth that these instruments can simplify malicious acts so simple as spear phishing, Thurai identified.

“Hackers are already utilizing these instruments to individualize content material, and it’s not simply ChatGPT,” he acknowledged. “One of many issues that you’ll be able to simply determine whenever you’re trying on the emails that are available in from a phishing assault is you take a look at among the key components in it, whether or not it’s human or automated AI content material. However these instruments have mastered the individualization of content material to imitate pure human habits.”

It isn’t all doom and gloom, nevertheless, as AI is already exhibiting the optimistic advantages it might deliver to cybersecurity, particularly in an age the place pattern-based malware scanning isn’t almost as efficient because it as soon as was, in response to Thurai.

“AI is nice at detecting issues which might be anomalies and decreasing the noise so you’ll be able to escalate solely the issues what you’re imagined to,” he stated. “AIOps is a superb use case to make use of within the safety areas, which they’re not utilizing to an extent but. Incident administration is one other space.”

Right here’s the entire video interview, a part of SiliconANGLE’s and theCUBE’s protection of CloudNativeSecurityCon:

Picture: SiliconANGLE