How One Cybersecurity Startup Knows Who’ll Be Breached

QuadMetrics is taking an offensive approach to fighting cybercrime.

The cybersecurity firm says it can predict the likelihood that a company will become the victim of a data breach with more than 90 percent accuracy, The Wall Street Journal reported Tuesday (Jan. 12).

By collecting and analyzing more than 250 various data points from its network, QuadMetrics uses its cloud service to identify the probability of a security breach taking place at an organization.

According to WSJ, the technology that QuadMetrics employs only requires external access to the company’s network to monitor its profile and compare it to predictive risk models. QuadMetrics then looks to historically reported incidents of data breaches and compares them to the cybersecurity characteristics of the company it is observing.

QuadMetrics CEO Wesley Huffstutter told WSJ that the process is similar to keeping an eye out on someone’s home from the street.

“We can notice that the windows are open, the doors are ajar or you only have one lock instead of two or your garage door is open,” he explained, while noting that other indicators, such as chipping paint, can help provide further details about what is going on.

“These aren’t necessarily security issues, but it goes to the fact that this may be a human element of whether you’re paying attention or not to what’s going on in your property,” Huffstutter said.

For each corporate structure examined, QuadMetrics assigns a risk score based on the models and gathered data about networks it uses. Huffstutter stated that the company has also identified the specific symptoms of network mismanagement that can lead to the accurate prediction of a security breach, adding that these elements are some of the most predictive capabilities within the company’s machine learning models.