English 
株式会社ジュピターテクノロジー株式会社エヌ・ティ・ティ・ドコモは、イタリア・レッコに本社を置くLibraesva Srl社が開発したメールセキュリティに特化したソリューション「Libraesva Email Security Gateway（ESG）」の販売を開始しました。
With the advent of generative AI, email phishing attacks have become more advanced and personalized. More sophisticated impersonation of colleagues, superiors, and business partners, natural-looking text that is indistinguishable from emails from overseas, and a variety of attack forms (not only traditional link clicks, but also attachments, deep fake audio/video, etc.) have become commonplace, and traditional rule-based security is insufficient, so the evolution of AI is necessary to counter threats. In addition, simply “using AI” is not enough, and there is an urgent need to utilize AI as a “purpose-specific solution.”
Libraesva Email Security Gateway (ESG) is an email security gateway that implements a unique AI engine called “Semantic AI” to significantly reduce the security risks of email users. Unlike “generative AI,” “Semantic AI” is a purpose-specific engine designed specifically to identify and classify email types, and its unique feature is that it can analyze the “intention of messages.”
こちらもお読みください： GMOサイバーセキュリティ、製造業向けCRAツールを提供開始
In addition, it uses a “small language model (SLM)” that is different from the commonly used large language model (LLM), and data is processed locally/edge and is not sent externally, including to Libraesva, reducing the risk of information leakage.
ソース PRタイムズ
日本語