AI in Cyber Security Part 1: Infiltration Into the Market

With ChatGPT capturing the attention of the global stage, dinner discussions and boardroom breakdowns alike are pondering the next ‘big thing’. Concern over generative AI is even creating widespread unease in the Hollywood writers’ scene.

In 2023, AI and ML (machine learning) are ubiquitous. We make artwork with it; we use it in ‘human-style’ chatbots and we’ve seen its ever-expanding spread into the business world. We even listen to music using it, thanks to Spotify’s latest ‘AI DJ’ offering.

In this 2-part blog we examine how AI is infiltrating the corporate cyber security landscape, from emerging threats powered by AI to enhancements in data security that might never be achieved without artificial intelligence or machine learning.

Emergence of AI in Cyber Security

Just as we saw the digital age heralded in by an explosion of devices and services, so too are we witnessing a gold rush of artificial intelligence opportunities. The inevitable threats aren’t too far behind them.

Chief among these initial discoveries is generative AI. Defined as an artificial intelligence that’s capable of creating media – think text, images and video – when fed a prompt.

Need a piece of art for your desktop wallpaper? A free AI will create it for you. Want a bedtime story written for your kids? It’s a question away. Looking for a brand-new piece of malicious code to conduct widespread, automated cybercrime with? Just type in a prompt; it’s probably even free.

Cutting-edge artificial intelligence solutions draw on the very best of AI and ML: the ability of both to act as force multipliers, empowering limited staff to manage data discovery & data classification, threat modelling & detection, email security and much more. Even behavioural analytics are drawing on the tech, with companies refining their User and Entity Behaviour Analytics solutions to help organisations determine organisational threats from user actions.

But what happens when AI gets in the hands of a malicious actor? Surely with AI thrown into the mix any malicious threat becomes even greater?

Even before AI’s emergence, the volume of cyber threats was alarmingly high. A 2021 cyber threat study conducted across North America, APAC and Europe deployed a range of honeypots to mimic likely targets of a cyber attach. The study found that 80% of honeypots were compromised within 24 hours. Staggeringly, all of the honeypots were compromised within a week.

We’ve all seen Terminator (although it’s unlikely we’ll witness AI taking over the world just yet). The realistic threat comes from cybercriminals pushing ahead in the arms race, drawing on the very resources and methodologies organisations are considering using to fight back.

Where benefits exist for organisations using AI, the opposite is true for criminals. As AI and ML offerings have flooded the open market – with many of them available for free – we’ve seen a corresponding hike in broader cybercrime statistics. In 2021, 86.2% of surveyed organisations were affected by a successful cyberattack.

Cyber security service providers worldwide are noticing marked increases in spam and phishing emails with mysteriously improved English-language grammar and spelling. You guessed it: Online AI platforms like ChatGPT and Google’s Bard are being used by scammers too. A reasonably well-written campaign is just a few clicks away for them now. According to a recent study, 9 in 10 data breaches are the result of phishing attacks.

Unleashing the Power of AI for Good Causes

Fortunately, AI and ML create improvements in technological use that we might never have experienced if artificial intelligence had not reached the heights it has today.

From better threat detection to improvements in process completion, AI and ML are already showing their strengths as some of the most practical and powerful applications of technology in modern cyber security. For example, well-trained algorithms can now rapidly detect any activity perceived as anomalous, with a level of sophistication in detection that typically trumps traditional methods.

To learn more, download our 16-page whitepaper on the emergence of AI in cyber security.