The Libraesva AI engine leverages a multilingual tensor-based model that builds upon core generative AI technology. Specifically designed for classification tasks, it provides tailored intelligence for distributed architectures, delivering high accuracy and performance.

The Libraesva AI Engine runs directly on our Email Security appliance, no data is exfiltrated. The AI engine focuses on analyzing emails from unusual correspondents or those exhibiting suspicious characteristics.

NOTE: A 20% average increase in CPU consumption can be anticipated.

The message scan result indicators, which provide a tag-based summary of complex analysis, have been enhanced to include both positive and negative signals and are now integrated with the new "AI classifier engine", enabling more accurate and insightful threat assessments.

Each indicator can also be used as a filter in saved searches and included in custom reports, allowing for more precise tracking and analysis.

Domain administrators can now configure their domain-based settings through the API. This enhancement allows for programmatic management of domain configurations, supporting automation and integration with third-party systems.