CITIC Telecom CPC: AI Visual Security

NOMINATION HIGHLIGHTS

SEEING Malware Quick and CATCHING Malware’s family FAST:
AI Visual Security is a game changer reinventing the “Seeing is Believing” security model, protecting enterprises from sophisticated malware with “Quick and Fast” trace. The out-of-box security solution is using AI-powered algorithm integrates weakly supervised regularization algorithm, visual computing, and neural network for transforming dataset into graphic image and mapping potential malwares, enabling enterprises to quickly ‘SEE’ malware and swiftly ‘CATCH’ malware families.

“Facial Recognition” for Malware:
“Hide and seek” malware creates “botnets” by quietly infecting massive numbers of devices using advanced communications. CITIC Telecom CPC created real noval breakthrough visual security solution in using computer vision with AI algorithms, transforming evasive malware bits and bytes data into 2D images, “mixed and matched” malware classification. CITIC Telecom CPC’s technique cleverly processes suspicious files with a 3D RGB color image algorithm. Then a specialized “autoencoder” and a “weakly supervised learning network” discover hidden features of suspicious files. Small visual images can represent the entire original data, and are easily managed. The amazing breakthrough is when the CITIC Telecom CPC method uses AI-enhanced computer vision to “view” transformed images, detecting malware. In other words, “AI Visual Security” is “facial recognition” for malware, successful even when malware “hides” using disguises. Additionally, this new method can use GPUs to handle image tasks, freeing up the main CPUs.This innovative solution can more efficiently and effectively tackle increasing, more frequent and complex threats, with thorough analysis of past threats and behaviors so as to prevent them in the future. AI Visual Security can also identify upcoming mutated threats because even mutated threats under the same family could still be easily identified before they pose any risk. Compared to traditional zero-day malware analysis using sandbox technology in the past, the AI Visual Security shows a significant improvement in the speed of identifying and classifying malware threats, estimated to be 10 to 100 times faster.