Magical Thinking Drives the Myth of AI Solving Security

SecurityNow, August 16, 2017

Every week, if not every day, a new security breach hits the news. Hackers continue to build their skills and launch increasingly sophisticated (and successful) attacks while security solutions are running out of breath just trying to keep up. Most traditional solutions simply can’t and never will catch up, even though their methods still remain in use despite little effectiveness. The latest area raising security hopes is AI – artificial intelligence driven by mathematics – intended to help computers detect and analyze patterns in high volumes. Advocates describe it as a magical, near “sci-fi” level of breakthrough. But can mathematics and pattern analysis really be the “end all, be all” for stopping the cleverness of the human brain?

Likely it can provide some measure of help in the broad spectrum. But at the same time, maybe there’s a reason magic is the stuff of myths and fairy tales more than practical reality. Would we could wave a magic wand here. But stopping human-sponsored, calculated file-less, memory-based attacks against web servers and applications are going to take more than past-focused signatures and machine-based algorithms. Security is still a realm where human brains must remain involved for solutions to be most effective.

Willy Leichter, vice president of marketing for Virsec, discusses specifics of what’s missing in the magical approach and what’s required to protect web servers and the apps that run on them.

Read full Magical Thinking Drives Myth of AI Solving Security article