Gone are the days when warfare was restricted strictly to armed conflict in physical space. Now, hackers and their targets can be seen in combat in the intangible digital landscape known as “cyberspace.” Simply stated, cyberspace refers to the virtual environment spanning multiple computer networks and interconnected machines on the internet. The pandemic brought with it an onslaught of cybersecurity breaches the world over, with Canadian and North American companies severely hit while having to make changes to their network systems in response to work-from-home models. This is partly due to the still nascent state of the global cybersecurity landscape and partly due to the sophistication of the aggressor’s methods.
In the most sophisticated attack ever reported of ransomware, Colonial Pipeline had to shut down its operations in May 2021 over a 5,500-mile network of gasoline-carrying pipelines connecting the Gulf Coast refineries to the eastern and southern part of the U.S.
Ransomware is a type of attack where malware (malicious software that can infect any computer system or network) is used to steal a company’s sensitive data and hold it for ransom in exchange for monetary demands. Attacks of this kind not only leak sensitive private data relating to consumers and company operations, but also bring a halt to operations, translating to millions of dollars lost in business and a blow to a company’s trustworthiness and reputation. According to a report by the cybersecurity company Emsisoft, in 2020, Canada experienced 4,000 ransomware attacks with the minimum ransom cost of almost 165 million dollars USD. With the inclusion of the losses incurred due to halted operations, the minimum loss figure leapt to a staggering estimate of over a billion dollars USD.
While another cybersecurity company, Check Point, reports that cyberattacks have increased by 93 per cent globally in mid-2021 from mid-2020, hope must still not be lost in the face of this boom in breaches, as artificial intelligence can be the much-needed saviour.
Why Artificial Intelligence in Cybersecurity?
Artificial intelligence (AI) refers to the design and engineering of computers that can think and act like humans. TechRepublic estimates the number of cybersecurity-related incidents that a medium-sized company needs to deal with on a daily basis to be 200,000, which is obviously nothing that a security team within the organization can tackle any given day. This is where AI comes in with the human-like ability to predict and respond to security threats. We will now look at more detail on how AI can transform the cybersecurity landscape.
Machine Learning Learns from the Past
A subset of AI, machine learning (ML) can be programmed to spot and record malware program information from previously held pools of data. When new malware is sent in attacks, ML systems can check against these records to search for similarities with malicious programs of the past to thwart the attacks. Attackers can and do try and outsmart the AI systems by concealing future attacks through slight variations in code. However, ML systems are strong enough to identify these disguises with their ability to recognize that the variants stem from a family of similar programs.
A very simple example of a machine learning system can be seen in Gmail, where the email software is designed to identify patterns of phishing emails that should end up in the spam folder. Most phishing emails attempt to gain private information such as passwords and banking credentials or hold links to ransomware. By analyzing email content, the context, and the metadata, the system can prevent phishing attacks from happening.
AI Identifies Vulnerabilities in Systems
A company’s network security is dependent on vulnerability management. Given that, on average, a medium-sized company faces threats numbering in several hundred thousand, detection, identification, and prevention of those is crucial. At a considerably much faster rate than human agency, AI can be leveraged to analyze the present tech infrastructure to identify the vulnerabilities and offer solutions. For example, consider this aspect of a vulnerability management system that can scour the social media and cybersecurity discussion forums for possible threats that others have had to deal with to tighten up the company’s security measures around those areas.
AI Responds Better in Real Time
A key feature of AI protection is its speed of identifying and responding to threats in real-time. Human intervention to detect threats and present timely solutions are severely limited. In comes AI with its ability to counter threats as they occur. As an example, consider McLaren Formula One’s deployment of AI by outsourcing the services of the cybersecurity fi rm Darktrace to constantly track data from cars, emails, chats, and many other sources to secure its network from any incoming threats. Moreover, the company can also make sure that it can work around the human tendency of cybersecurity personnel to seek shortcuts at the job, thereby compromising the security of its system.
AI Can Search for and Install Security Updates
Even for an average medium-sized company, its internal network of computers and other connected devices can be quite expansive or based over several geographic locations. Another important way in which AI can help with the maintenance of security is by checking which devices or software needs updated security patches or other updates that will improve security across the entire network. Most often for hackers, buggy software to which security patches haven’t been applied becomes a prime candidate capable of exploitation.
While AI may come across as a magical solution to protect against cybercrime, its efficacy is dependent entirely on sensible utilization. After all, the agents behind any AI engine are humans, and human intervention is certainly necessary for any security system. It must be remembered that cybercriminals can also deploy AI to predict solutions and devise more complex malware to derail a company. Hence, cybersecurity personnel must know how to optimise the AI code to pre-emptively thwart counterattacks.
Arslan Ahmed | Staff Writer