Trend Micro is making the case that cybersecurity has reached a level of complexity that requires increased reliance on machine learning algorithms and other forms of artificial intelligence (AI)—technologies it will manage on behalf of customers when required.
Steve Neville, director of global strategy for Trend Micro, said the Trend Micro Managed Detection and Response (MDR) service can aggregate data from more than 250 million endpoints to enable machine learning algorithms to prioritize alerts and identify high-risk threats faster than cybersecurity teams could on their own. Armed with that information, cybersecurity teams can orchestrate and automate the appropriate levels of response.
Machine learning algorithms will be embedded across the entire Trend Micro portfolio, Neville said, so customers can either opt to employ a managed service provided by Trend Micro or deploy the company’s security software on their own.
Neville said Trend Micro MDR is intended to augment cybersecurity professionals, not replace them. In some cases users will access Trend Micro MDR directly, while in other cases Trend Micro MDR will complement the capabilities of managed security service providers (MSSPs).
Neville said AI isn’t a “silver bullet” for cybersecurity. But it does enable cybersecurity experts, who are in short supply, to cope with cyberattacks that are increasing in volume and sophistication.
Most organizations aren’t able to aggregate the volume of cybersecurity data required to train machine learning algorithms to recognize and prioritize threats, so for all intents and purposes adoption of AI technologies to combat cybercriminals will need to be driven by a vendor that can collect data from millions of sources. Over time, those capabilities should significantly improve the ability of cybersecurity teams to respond to new threats as well as hunt for threats that have made it past their primary cybersecurity defenses.
Given the chronic shortage of cybersecurity professionals, the rise of AI should not result in any loss of cybersecurity jobs. But the nature of those jobs likely will change, as many rote tasks become increasingly automated by machines.
Neville said that at this point there are no indications that cybercriminals have begun to employ machine learning algorithms and other AI techniques to better target their attacks. But, given the amount of financial resources cybercriminals have at their disposal, it might only be a matter of time before they do, he said.
Cybersecurity professionals should assume that something akin to an AI arms race is now underway. After all, most of the machine learning algorithms being used to drive AI models have been in the public domain for decades. The good news is that—for once, at least—cybersecurity vendors might be temporarily ahead of cybercriminals in terms of being able to take advantage of machine and next-generation deep learning algorithms. How long that advantage might last, however, is still anybody’s guess.