日本語 
RainForest Inc. is pleased to announce that it has officially implemented an enhanced CPE automatic identification engine function that utilizes generative AI (LLM: Large Scale Language Model) in its domestically developed cybersecurity products “Senda Series (Senda-Discover and Senda-Radar)”. In addition to the conventional signature-based system, the AI performs identification and estimation processing for unknown banner information and spelling variations, significantly improving the accuracy of visualization of the attack surface and the comprehensiveness of vulnerabilities.
In recent years, as the importance of visualization and control of the attack surface has increased, there is a demand for automation of service identification and vulnerability management in addition to traditional port scanning and banner analysis. RainForest has been researching and developing the collection, visualization, and analysis of risk information on external and internal networks through two products, Senda-Discover and Senda-Radar.
Also Read: NaLaLys Secures ¥250M in Pre-Series A for AI Compliance
The CPE identification engine common to these products uses a mechanism to infer the service name and version from the banner information and check it against vulnerability databases such as NVD. However, actual banner expressions are often unstructured and have variations in description, so there were cases where detection misses and false positives were unavoidable.
SOURCE: PRTimes
English