MARCH 1, 2016

Deep Instinct Files Portfolio of Patents to Implement Deep Learning in Cybersecurity

Patent-pending technology applying deep learning to cybersecurity for the first time prevents zero-day and APT attacks on endpoints, servers and

Patent-pending technology applying deep learning to cybersecurity for the first time prevents zero-day and APT attacks on endpoints, servers and mobile devices, freeing enterprises from costly recovery

 SAN FRANCISCO--(BUSINESS WIRE)--Today, Deep Instinct announced that it has filed five patents covering the only technology that applies deep learning to cybersecurity. The extensive patent applications, which contain, in the aggregate, over 150 independent claims, cover the entire innovative body of knowledge achieved by Deep Instinct in deep learning and cybersecurity: both for currently commercially available products and for future planned products.

Deep learning is a novel branch of artificial intelligence. It is considered by many researchers in the field of computational intelligence to contain the most suitable family of algorithms for domains that require the ability to analyze vast and complex data, such as the data comprising files in cybersecurity.

Deep Instinct is the first company applying deep learning to detect, in real-time, malware on endpoints, servers and mobile devices, focusing on zero days and APT attacks – areas where traditional cybersecurity practices lack the capacity to protect in real-time.

In order to obtain a trained deep learning model, the neural networks comprising the deep learning must be trained on a very large corpus of data. Following its training phase, the deep learning model can then operate in a prediction mode. Applying deep learning to both phases – training and prediction – required Deep Instinct to develop tailored methodologies to optimize the detection rate, as well as other relevant parameters, such as speed and consumption of resources. The patent applications, which are confidential, utilize deep learning libraries and algorithms that were entirely created by Deep Instinct, and cover the novelty of Deep Instinct’s technology in these operational phases of the deep learning model. They relate to Deep Instinct’s ability to condense its deep learning model into a small, light agent that is installed on any endpoint, server and mobile device and operates in an autonomous manner to detect and prevent any type of malware.

Deep learning facilitated further ingenuity that led to filing one of the five patents that pertains to traffic. In an unsupervised form, deep learning is the best methodology for identifying anomalies, minor mutations, and non-regular activities. These are areas where traditional cybersecurity approaches lack the capacity to provide comprehensive, real-time protection. Deep Instinct has claimed patent protection with respect to its novel capability to apply deep learning to Deep Packet Inspection (DPI) and Packet Analysis to detect these phenomena in the network of any enterprise or other entities (including smart vehicles) from raw data algorithms.

“The remarkable technological capabilities of deep learning and our particular expertise enabled us to create an extensive portfolio of unique patents in cybersecurity, which to the best of my knowledge, make Deep Instinct the first company to file such a broad range of this type of patents.” said Guy Caspi, CEO of Deep Instinct. “I am proud to share this complex and lengthy process that we have undergone, which spanned over a year, starting in 2015 and ending when we made the patent applications.”

Please visit Deep Instinct at RSA booth 4903 North.

About Deep Instinct

Deep Instinct is the first company to apply deep learning to cybersecurity. Leveraging deep learning’s predictive capabilities, Deep Instinct’s on-device, proactive solution protects against zero-day threats and APT attacks with unmatched accuracy. Deep Instinct provides comprehensive defense that is designed to protect against the most evasive unknown malware in real-time, across an organization’s endpoints, servers, and mobile devices. Deep learning’s capabilities of identifying malware from any data source results in comprehensive protection on any device, any platform, and operating system. Deep Instinct is headquartered in Tel Aviv, Israel and has offices in San Francisco, CA.

To learn more, please visit


Deep Instinct
Maya Schirmann, +972 (3) 545 6600
Vantage PR for Deep Instinct
Lynda Starr, +1 973-386-5949