Are Applications of AI in Cybersecurity Delivering What They Promised?

February 12, 2019
| |
3 min read

Many enterprises are using artificial intelligence (AI) technologies as part of their overall security strategy, but results are mixed on the post-deployment usefulness of AI in cybersecurity settings.

This trend is supported by a new white paper from Osterman Research titled “The State of AI in Cybersecurity: The Benefits, Limitations and Evolving Questions.” According to the study, which included responses from 400 organizations with more than 1,000 employees, 73 percent of organizations have implemented security products that incorporate at least some level of AI.

However, 46 percent agree that rules creation and implementation are burdensome, and 25 percent said they do not plan to implement additional AI-enabled security solutions in the future. These findings may indicate that AI is still in the early stages of practical use and its true potential is still to come.

How Effective Is AI in Cybersecurity?

“Any ITDM should approach AI for security very cautiously,” said Steve Tcherchian, chief information security officer (CISO) and director of product at XYPRO Technology. “There are a multitude of security vendors who tout AI capabilities. These make for great presentations, marketing materials and conversations filled with buzz words, but when the rubber meets the road, the advancement in technology just isn’t there in 2019 yet.”

The marketing Tcherchian refers to has certainly drummed up considerable attention, but AI may not yet be delivering enough when it comes to measurable results for security. Respondents to the Osterman Research study noted that the AI technologies they have in place do not help mitigate many of the threats faced by enterprise security teams, including zero-day and advanced threats.

Still Work to Do, but Promise for the Future

While applications of artificial intelligence must still mature for businesses to realize their full benefits, many in the industry still feel the technology offers promise for a variety of applications, such as improving the speed of processing alerts.

“AI has a great potential because security is a moving target, and fixed rule set models will always be evaded as hackers are modifying their attacks,” said Marty Puranik, CEO of Atlantic.Net. “If you have a device that can learn and adapt to new forms of attacks, it will be able to at least keep up with newer types of threats.”

Research from the Ponemon Institute predicted several benefits of AI use, including cost-savings, lower likelihood of data breaches and productivity enhancements. The research found that businesses spent on average around $3 million fighting exploits without AI in place. Those who have AI technology deployed spent an average of $814,873 on the same threats, a savings of more than $2 million.

Help for Overextended Security Teams

AI is also being considered as a potential point of relief for the cybersecurity skills shortage. Many organizations are pinched to find the help they need in security, with Cybersecurity Ventures predicting the skills shortage will increase to 3.5 million unfilled cybersecurity positions by 2021.

AI can help security teams increase efficiency by quickly making sense of all the noise from alerts. This could prove to be invaluable because at least 64 percent of alerts per day are not investigated, according to Enterprise Management Associates (EMA). AI, in tandem with meaningful analytics, can help determine which alerts analysts should investigate and discern valuable information about what is worth prioritizing, freeing security staff to focus on other, more critical tasks.

“It promises great improvements in cybersecurity-related operations, as AI releases security engineers from the necessity to perform repetitive manual processes and provides them with an opportunity and time to improve their skills, learn how to use new tools, technologies,” said Uladzislau Murashka, a certified ethical hacker (CEH) at ScienceSoft.

Note that while AI offers the potential for quicker, more efficient handling of alerts, human intervention will continue to be critical. Applications of artificial intelligence will not replace humans on the security team anytime soon.

Paving an Intelligent Path Forward

It’s important to consider another group that is investing in AI technology and using it for financial gains: cybercriminals. Along with enterprise security managers, those who make a living by exploiting sensitive data also understand the potential AI has for the future. It will be interesting to see how these capabilities play out in the future cat-and-mouse game of cybersecurity.

While AI in cybersecurity is still in the early stages of its evolution, its potential has yet to be fully realized. As security teams continue to invest in and develop AI technologies, these capabilities will someday be an integral part of cyberdefense.

Download the ebook: 7 questions to ask before adopting AI in your SOC
Joan Goodchild
Contributor

Joan is an award-winning veteran journalist, editor, writer, researcher. She is a seasoned correspondent covering the security industry with deep contacts an...
read more