Page 257 - Cyber Defense eMagazine Annual RSA Edition for 2024

How Is AI Used in Cybersecurity?

Nowadays, cyber threats have gotten smarter, significantly increasing ever since people started working
remotely or in a hybrid system. Hackers are finding more ingenious ways to breach even the strongest
systems, leading to losses worth billions of dollars. Traditional efforts from the past have been effective
in keeping the treats at bay, but with their intensity increasing, they are no longer enough.

As a result, more and more companies have been employing AI to analyze significant amounts of
information within a short amount of time. Its popularity is due to the ability to spot suspicious patterns
and abnormalities, identifying an attack before it can pass through the firewalls. The speed is often faster
than human intelligence can perform, leading to quick resolution strategies.

On the other hand, just like major companies are using AI to protect themselves, hackers have learned
how to harness its power for complex attacks. Generative AI has been responsible for supporting
numerous phishing attacks, malware, ransomware, and even insider threats. This led to a need for even
stronger security protocols, protecting against the same system that supports their own protection.

Cybersecurity – A Benefit Rather than a Threat

Even though many are still wary of artificial intelligence’s effect on their jobs, an average of 78% of
company leaders admit to using generative AI. The reasoning behind that is relatively simple: its data
poll is much larger than what human attention can contain. As a result, its long-term use will likely bring
the following advantages:

Automatic Detection of Vulnerabilities
Potential vulnerabilities are very often missed when human-driven security methods are employed. This
can lead to the release of faulty, unsecured applications that can put users and company owners at risk.
Artificial intelligence can examine applications for weak spots, reporting incorrect configurations. While
AI technology is still not reliable enough to fully handle the correction itself, it can raise the alarm and
showcase the problem. This will help prevent DDoS attacks and other cybersecurity threats.

AI-on-AI Attack Detection

Numerous AI detection tools have already been created to determine whether or not another project was
finalized using artificial intelligence. As both types of AI tools are likely tapping into the same data pools,
it’s easy to guess what was written by its ‘peer’ and what was not. The same ability is expected to
eventually reach the cybersecurity world, where companies can use AI to detect other attacks fueled by
the technology. This can significantly protect future businesses from a potential breach.

257