AI & ML: Safety or Threat for Cyber Security


The Universal Health Service (UHS) attack that happened recently has attracted the crowd’s attention to the threat of ransomware in general and health systems in particular. These attacks are one of the reasons why AI and Machine Learning are forced to have a bad image.

The current COVID-19 situation has opened up a tremendous opportunity for cybercriminals. The newly launched telephone programs, IT deployments, work from home models, sophisticated and fear-based phishing techniques make up many firms as easy targets. As these took over, many industries installed untested platforms and the lack of training are also the reasons. So what can be done to shield against such attacks?

Malware doesn’t ask for much user interference once access to a system is gained. The same AI and ML can be used to compromise as well as to defend systems’ security which is why experts say that these can be a double-edged sword. Also, these can contribute in multiple ways to security that are yet to be explored. One of these includes the user and entity behavior analytics that study an individual’s distinctive behavior and flags any abnormal behavior.

Another application of AI and ML is to spot subtle behavior patterns among attackers. Phishing emails can often be a would-be victim’s emotions such as playing up the urgency to compel someone to click on a link. The sentiment analysis can be used to spot such emails or if it’s content seems abnormally heated. The email structures can also be used to spot the attackers as they may rely on templates or available structure to provoke responses.

However, this doesn’t mean that AI is perfect against attacks, it is also subjected to bias as its creators. These algorithms work very delicately which allows them to be corrupted just like any other technology. The email filtering solutions, layered security controls, network visibility, and data control assist to ensure system safety.

The applications of AI and ML are just the beginning for now and more areas are yet to be explored. Until then human engineering is to be given importance. This involves training employees on implementing strong security check and to recognize and flag suspicious behavior.


Please enter your comment!
Please enter your name here