How Artificial Intelligence and Machine Learning Can Redefine Cybersecurity

Cyber attacks have been rapidly increasing in both complexity and volume over the last few years. Vulnerabilities in our critical infrastructure have already been exposed, and the virtual game of cat and mouse with cybercriminals is seemingly just getting started. But where do we begin to protect our essential services, personal data, and businesses?

The inconvenient truth that 90% of all successful cyber attacks are attributable to either human error or inaction suggests that education is more important than any new shiny technology solution. But maybe we can all agree that a proactive rather than reactive approach to cybersecurity should be top of everyone's agenda.

According to Cybersecurity Ventures, cybercrime could reach $6 trillion by 2021. Once again, a combination of Artificial Intelligence and machine learning will play an important role both in cybersecurity and as the weapon of choice for the cybercriminals.

For these reasons, AI will be seen as a double-edged sword in the world of cybersecurity. Sure, businesses will rush to embrace the additional protection and benefits that this technology can bring. On the other hand, the bad guys will be looking to use it to automate attacks and prove that it's our fellow humans rather than technology that represent our biggest threat.

It's important to highlight how the threats are not just in the form of data breaches or about bringing large corporations to their knees. In industries such as aviation, a cyber attack can be devastating. For example, the 2008 Spanair crash that killed 54 people was the result of a Trojan horse that infected the central computer system used for monitoring technical problems in airplanes.

Elsewhere, Ukraine suffered a significant cyber attack on a power grid that left thousands of homes without any power. Meanwhile, the infamous WannaCry ransomware attack affected more 10,000 organizations in 150 countries and the national health service (NHS) hospitals in the UK.

However, machine-learning models are being deployed to help companies better protect themselves against these sophisticated threats. When armed with the same tools as the criminals, it also becomes easier to scale or learn patterns and identify deviations in ways that humans alone could never manage.

Thankfully, help is at hand. MIT’s Computer Science and Artificial Intelligence Lab currently developing a cybersecurity platform that will make it easier for analysts to review data and flag security threats.

With the entire critical infrastructure at risk, it should be of little surprise that the U.S. government is one of the biggest players in the AI space for cybersecurity. Predictably, they are designing AI-based software that can detect and respond to threats quicker than is humanly possible, and it plays a crucial part of our national defense.

Advances in AI technology are enabling analysts to move away from a reactive approach where they were reliant on prior knowledge of the threat type before they were able to defend from it. Alternatively, machine learning is much more proactive by continuously learning, adapting, and improving to preempt the previously unseen attacks.

However, AI doesn't just speed up responses or make it easier to stop potential attacks before they occur. The really exciting aspect of this technology is how it can also recognize changes in an attack methodology and protect accordingly.

AI and machine learning are a perfect team and crucial in the fight against attackers with their ability to analyze vast amounts of data quickly and comprehensively. However, they are unable to replace human decision-making when protecting corporations from harm. Once again, this highlights how AI works best when it works alongside humans, rather than replacing them.

The most significant challenge is that our human adversaries will be armed with a smorgasbord of motives. They will also be using the same technology to counteract cyber defenses to attack networks undetected. This modern game of chess between two teams that represent both the best and worst of humanity and technology comes with an element of irony.

One of the greatest threats to society and indeed our planet is our own free will. It's humans who program and design machines using the latest technology, and some will use it to make a difference, while others will try and leverage AI for nefarious activities. The battle for the greater good is not about pitting humans vs. machines; it's about pitting humans against each other, and this is probably the saddest part of any future cyber war.

Only time will tell if AI and machine learning will make a positive impact in the cybersecurity industry. Although our future world looks certain to be different from the one we inhabit today, unfortunately, it seems that many of the flaws in the human character will continue to hold back real progress.