Machine Learning: Test, Test, Test Some More

Seventy percent of security researchers said attackers can bypass machine learning-driven security technologies and nearly one-third – 30 percent – said those solutions are easy to bypass.

Machine learning "is a critical component to next-generation endpoint security technologies," Carbon Black said in the report. But the fault with the technology lies in "how much emphasis organizations may be placing on it and how they are using it."

Rather, the most resilient ML approaches involve dynamic analysis - evaluating programs based on the actions they take.

What should businesses do to root out malicious behavior in machine-learning technologies? Carbon Black recommends that they employ a "massive body of baseline data, a torrent of detonation data," and statistics and comparisons among behaviors for validation. "Collectively, these approaches will give you a powerful set of tools to generate patterns of malicious behavior," the report read.