How AI can be the next leap for cybersecurity

Image of padlock against circuit board/cybersecurity background
(Image credit: Future)

With the growing threat of cyberattacks, AI is the answer to rebalance the cybersecurity game and move the industry forward. Cybersecurity historically has evolved out of necessity as threats emerge and existing solutions become ineffective. Since the first example of antivirus software was developed, new technologies and subsequent vulnerabilities require innovation to combat them. This is why AI is essential in today’s fight against cybercriminals.

About the author

Andrew B. Gardner is VP of Research and AI at Avast.

Constant innovation in cyber threats and protection

When new virus and malware numbers exploded in the 1990s, it became clear that cybersecurity had to be mass-produced to protect the public. By 1996, many viruses used new techniques and innovative methods, including stealth capability, polymorphism, and ‘macro viruses’ – posing a new set of challenges for antivirus vendors.

As Internet access became widely available throughout the 2000s, increasing connectedness and ongoing digitization afforded cybercriminals more opportunities than ever before. The continual discovery of zero-day attacks combined with an exponentially larger user base meant that antivirus was becoming less effective. By the 2010s, the world began to see numerous high-profile attacks impacting the national security of nations and costing businesses millions.

The cybersecurity industry is now bigger than ever and has evolved to tackle an expanding range of attack types with new approaches to protection, such as Multi-Factor Authentication (MFA), real-time protection, Sandboxing and Network Behavioral Analysis (NBA). In turn, criminals responded with their own innovations: multi-vector attacks and sophisticated social engineering. As attackers become smarter, antivirus has been forced to shift away from signature-based detection methods to ‘next generation’ innovations that use different approaches to increase detection of new and unprecedented threats. These threats go far beyond device hacking, with cyber now a key battleground, national priority for countries like the USA and, increasingly, the source of international political tensions.

To defend against the new landscape, cybersecurity developers must look to the cutting edge of technology to modernize their work and keep up with bad actors.

How AI can rebalance the computer security game

AI is one of the newest developments in security, modernizing previous approaches by disrupting the industry and helping us to think alternatively. Computer security used to simply be antivirus on the computer, but these days the key devices that need protection are smartphones, tablets, and other IoT devices. Cybersecurity is now more about making sense of how we interact, where we place trust and where we spend our attention. Our interactions with other devices and other people are also amplified by social media, e-commerce, and digital transformation in our daily lives.

Artificial intelligence has huge potential to transform cybersecurity. In fact, with an ever-growing deployment of AI-based automation by attackers, the volume of relevant machine data is growing so quickly that using modern machine learning techniques is the only way that we can provide cybersecurity today. AI can help by protecting us from clear threats and less clear threats across these new interactions without requiring explicit direction. We need to minimize the dependence of defense on human experts who are simply not equipped to combat AI-assisted attacks. Automated feature extraction is the proper answer to such a challenge, and it provides expert security analysts with the ability to focus on the most sophisticated attacks created by human attackers.

Sophisticated threat prevention in today’s world consists of a combination of multiple machine learning engines that work together to defend against attacks. Avast has built a unique and sophisticated machine learning pipeline that allows us to rapidly train and deploy malware detection models within hours. We also employ advanced techniques like deep convolutional neural networks to enhance our malware detection models. New security threats can appear suddenly and take new and unknown forms; in such situations, the ability to update models rapidly ensures users remain protected. This next-gen security technology, and the data from Avast’s massive user base, gives a clear advantage against hackers.

The future of AI in cybersecurity

I have huge hope for the future of AI in cybersecurity and, more broadly, in the public domain. I hope to see the industry move forward and shape the future of computer security by moving far beyond the traditional idea of protection from file, script and email threats to systems that protect transactions, interactions, conversations, and attention.

If I could set a goal for Avast and our industry, it would be to revise how people view security products and AI. AI is so often misunderstood, typically because of the lack of universal definition, which leads to confusion and ultimately disappointment. There is a real understanding gap that requires addressing because AI’s impact on society is so complex. Fundamentally, it is an intelligent system that can collect and process data, but decision-making is the real hallmark of AI.

AI developments in cybersecurity will ultimately secure our digital future, and with increasing public understanding, I hope to see users trust and engage with AI in security products. The role of AI in cybersecurity is now essential to enabling people to enjoy digital freedom, and I look forward to seeing the development of new AI-based solutions which make the most of its huge potential.

Andrew B. Gardner is VP of Research and AI at Avast.