Leveraging Artificial Intelligence & GPUs for Cybersecurity

This post is sponsored by NVIDIA. All thoughts and opinions are my own. 

Artificial Intelligence (AI) presents a significant opportunity to solve problems previously either not easy to solve or worse, not possible to solve. The combination of AI along with today’s Graphics Processing Unit (GPU) technology provides an added boost to those leveraging sophisticated algorithms in their deep learning solutions. These sophisticated systems are able to train deep learning models and ultimately lead to predictive insights. The objective is to move from reactive to proactive and finally to predictive insights.

The breadth of opportunities that AI presents is wide, however, a significant opportunity is in the Cybersecurity space. One company leveraging the power of AI & GPUs is recent Nvidia Inception Program award winner Deep Instinct. Deep Instinct leverages deep learning to provide insights for zero-day malware detection in both endpoint and mobile devices; two of the key areas of concern in cybersecurity.

The range of today’s cybersecurity threats present a bevy of challenges. The challenge is with complexity and speed; two characteristics that often work against each. Yet, the combination of AI & deep learning provides the foundation to bring speed to solving complex problems like those in the cybersecurity space.

This powerful combination moves the paradigm from reactive to predictive solutions and can mean the difference between breach and prevention. As zero-day threats become more frequent and damaging, only through the use of sophisticated AI models can one have the potential to mitigate these risks.

Fraud detection is yet another area in the cybersecurity space that is gaining attention. Detecting fraud in real-time has typically been done based on policies set by humans. Yet both of these risks (zero-day and fraud detection) are evolving faster than humans can keep up. We need a better approach.

Solving complicated problems is not easy. One must always be one step ahead of the bad actors. As we move to a paradigm where security is part of our DNA, so must the approach we take. We can no longer afford to be reactive. Only the proactive…and predictive will remain relevant.

Tim Crawford is Gigaom Head of Research, and strategic CIO and consultant. Read Bio »