Posted by Though AI in Cybersecurity is still evolving, it is already becoming a new conundrum for CISOs and security professionals. When almost everyone (including the cybercriminals) and everything today has started to leverage the power of AI, shouldn’t the CISO evaluate Artificial Intelligence in Cybersecurity to enhance the capabilities of security tools and solutions before it’s too late? Yes, today, that’s the new conundrum that CISOs are dealing with.
A CISO’s job is to safeguard and protect his/her organisation against cyberattacks and thwart data breaches by hackers and cybercriminals. But when these cyber crooks have already started to leverage AI to hit their targets with more powerful and effectiveness, can the CISO afford to overlook Artificial Intelligence in Cybersecurity in his/her organisation? The answer is a big no.
AI in Cybersecurity
Having said that. For CISO it’s not that simple and easy to take a call on how to leverage or deploy AI in the Cybersecurity unit that will help to enhance or augment capabilities and cyber defence.
So how should CISOs go about dealing with the new conundrum of leveraging AI in Cybersecurity or incorporating AI into the Cybersecurity strategy of the organisation?
“I recommend approaching AI like you would a candidate for a job on your team. Assess its effectiveness, ease of use and trustworthiness,” Check Point Software Technologies CISO Jony Fischbein wrote in a blog post.
In the post, CISO Fischbein has posed three key questions that every CISO should ask while assessing AI in Cybersecurity.
How is AI being used to augment cybersecurity capabilities?
The first question he put forth to fellow CISOs.
“One of AI’s benefits is its creativity and ability to make previously unheard of – yet genius – decisions,” CISO Fischbein wrote in his reply. To explain his answer, CISO Fischbein cited the example of Google acquired-London-based DeepMind Technologies’ AlphaGo, a computer program that plays the board game Go, which had defeated the reigning Go world champion, Lee Sedol in 2016.
“Go is an ancient and exceedingly complex strategy game. During the match, AlphaGo made a move that confused Go experts, who thought it was a strange mistake. But Move 37, as it came to be known, was actually the turning point for the match – and one which Sedol wasn’t able to overcome. It’s not a move a human would have made,” CISO Fischbein explained.
Based on AlphaGo’s example, he asked fellow CISOs to look for a solution that uses AI to prevent threats that other vendors can’t even yet detect. “Ask about their innovation cycle and what threats they see on the horizon,” he added.
What is the level of AI expertise?
The second question he asked.
“With the current popularity of AI, many companies are rushing to add some level of AI capabilities to their products. But in this economy, CISOs are being asked to run operations more efficiently and need to justify budgets,” CISO Fischbein wrote in his reply to his second question.
“There’s no need to pay for limited AI capabilities. Ask for third-party validation of their AI solution’s accuracy to determine whether they are providing real value or simply creating more noise and false alerts,” he added.
Probably, the first question may sound broader in nature for some CISOs, but certainly, the second question is more likely to give precise data points in evaluating and making decisions in investing in AI.
CISO Fischbein’s third question is more intriguing and interesting at a time when almost everything claims to have AI in some form or capability. Third question he asked.
Can AI technologies be trusted?
In the answer, he wrote that “AI models” are only as “good” as the “quality and quantity of the data” they are trained with. That’s the Gospel truth about the AI model and it’s an open secret that most vendors and their marketing and sales teams are well aware of. And yet, they would make tall claims about their products and mint money from companies and businesses.
To make his point, he quoted Stanford Professor James Zou, “One of the best ways to improve algorithms’ trustworthiness is to improve the data that goes into training and evaluating the algorithm.”
“Look for a solution that provides real-time threat updates and has a large customer base. The more customers, the more training data available for the AI,” he added.
“With the rate and sophistication of cyberattacks increasing every year, as CISOs, we need every advantage we can find to protect our data and teams. AI may offer a powerful advantage, as long as we are deploying trusted solutions that move beyond hype to reality,” CISO Fischbein concluded his blog post.
Certainly, the three questions by CISO Fischbein can help CISOs to better under AI and its capabilities before they decide to implement or invest in AI as part of AI in Cybersecurity strategy or approach in their respective organisations.