Dubai: How generative AI is playing key role in future of cybersecurity

In an ever-evolving digital landscape, companies are spending copious amounts of both time and money to keep themselves safe. Whether it’s dealing with phishing emails, network breaches, or any other myriad of digital threats, the race to stay one step ahead of attackers is well and truly on. Among the variety of tools and services being employed to fortify enterprise networks, one surprising ally has emerged – Artificial Intelligence.

While the landscape for its application is vast, AI has a very specific role to play when it comes to cybersecurity. AI excels in tasks that require repetition or pattern recognition, which makes it ideal for detecting attack patterns and deploying safeguards autonomously. “Techniques like DDoS mitigation and bot detection are AI-enabled and help mitigate security threats right where they start from. For instance, an AI could automatically isolate affected systems that contain ransomware, or deploy patches to fix vulnerabilities, reacting much faster than traditional security measures,” commented Ramprakash Ramamoorthy, Director of AI research, ManageEngine.

Narrowing down the field further, Generative AI is the tool to focus on, and companies that are able to understand and harness its capabilities will remain miles ahead of the potential attacks on their digital infrastructure. “One of the core challenges in cybersecurity is the scarcity of skilled professionals. The fast-evolving threat landscape demands a workforce that is both adept and numerous. Generative AI serves as a pivotal force multiplier in this regard, helping to bridge the workforce gap by enhancing the capabilities of existing digital systems and staff,” remarks Meriam El Ouazzani, Regional Director, Middle East, Turkey, Africa, at SentinelOne.

Artificial Intelligence is still a very new skill to understand and master accordingly, so until the market holds more talent with this particular skillset, Generative AI can help fill in the gaps.

Of course, an AI is only as good as the data it receives, so it is imperative that the data being supplied to these AI systems is relevant to the system it is trying to defend. “The challenge in developing a good AI model is figuring out how to get your model to detect threats that are currently not being detected (and hence labelled) by your existing detectors, otherwise you risk the model just being a very expensive parrot of your existing detection rules,” says Harish Chib, Vice President, Middle East & Africa, Sophos.

• ‘Super iPaaS’ to manage growing chaos of connectivity