Can AI Design Your Cyber Defenses?

 

Carl Mazzanti is the president of eMazzanti Technologies in Hoboken.

We are living through a period where artificial intelligence (AI) is seamlessly woven into the fabric of our daily lives, from seeking directions on our cell phones to Netflix suggesting the next binge-worthy show. The spotlight is now shifting to a critical concern: who holds the reins when it comes to Cyber Security measures? Trust becomes the linchpin in this debate and, unsurprisingly, AI takes center stage in the controversy.

The essence of Cyber Security lies in the relentless pursuit of threats and the detection of anomalies. If something does not “smell” right, it should be flagged and thoroughly investigated. But as the efficiency and frequency of cyberattacks surge exponentially – with reports indicating that the average business faces over 1,000 attacks weekly – the question is: who will spearhead the effort to combat this deluge of threats?  

Cyber Security Solutions and AI

Recent technological strides suggest that the answer lies in AI. Microsoft, for instance, has already integrated AI into evaluating Cyber Security solutions. They have even gamified the machine learning process, turning it into a virtual battleground where attackers attempt to steal confidential information, defenders strive to thwart them, and the ultimate prize is winning the game.  

Yet, as AI evolves to the point of assessing a company’s security environment and recommending configurations and maintenance, entrepreneurs must confront some issues: if a tool can conduct an audit, should it be entrusted to carry out the entire job? This question today will shape the landscape of Cyber Security shortly.  

Consider this example: a business opts for multifactor authentication (MFA), demanding users provide two or more verification factors for access. MFA is mandated on all devices to ensure stringent security. However, exceptions may arise a sudden need for access or a device that requires immediate use without proper configuration. Right now, a human decides on these exceptions. But what if a cascade of exceptions — each seemingly reasonable on its own — results and jeopardizes the company’s Cyber Security? What if an automated cybersecurity tool highlights weaknesses, but the recommendations are ignored?  

These possibilities lean towards allowing AI to take the lead in Cyber Security. But caution flags still need to be raised. What if an AI program is inadvertently infused with common biases that could permeate every decision, rendering multiple companies vulnerable across various industries?  

What is the Solution?

I believe the solution does not hinge on an all-or-nothing approach, like the one portrayed in The Matrix – where the choice was surrendering all control to AI or simply pulling the plug. Instead, gaining a deeper understanding of how AI arrives at its decisions and striking a balance is imperative. The demand for swift Cyber Security solutions is escalating, and meeting this demand involves more integration of AI. But humans must retain the ultimate responsibility for scrutinizing AI decisions rather than blindly accepting them.  

The challenge, then, falls on Cyber Security providers to cultivate and elevate data scientists and other professionals who are capable of peering beyond the veil of AI and can comprehend its decision-making processes. Just as businesses embracing AI are poised to replace those who resist, Cyber Security providers equipped with an understanding of AI’s operations are destined to supersede those who lag.  

In essence, the future of Cyber Security lies in a symbiotic relationship between human oversight and AI capabilities. As we navigate this evolving landscape, the key is not to relinquish control, but to wield it judiciously, ensuring that the union of human intuition and artificial intelligence further fortifies our digital defenses against the rising tide of cyber threats.