The conversion process will be performed automatically during installation and may take several minutes to complete, depending on the size of the database and the performance of the disk where it is stored. During this time, SecurityGateway will be unavailable.
Before the update, a backup of the database file will be created. The backup file, named SecurityGateway.fb3, will be stored in the SecurityGateway\App directory.
Please note that once the database file has been updated, it will no longer be compatible with earlier versions of SecurityGateway.
Administrators are able to configure AI models from various providers including OpenAI/ChatGPT, Google Gemini, and custom API endpoints. The system allows the creation of custom AI prompts using message data variables that ask the AI to classify messages into administrator defined categories. Classification rules can be used to trigger specific actions based on the AI's classification results.
Any model that supports the OpenAI API format can be used, including local models running on your own infrastructure. Users are responsible for obtaining any necessary API keys and for any costs incurred when using third-party AI services.
AI Classification provides an advanced method to identify sophisticated phishing attempts, detect sensitive content such as personally identifiable information (PII), and filter unwanted messages that traditional rules might miss.
This approach is more secure and allows the use of Microsoft 365 two-factor authentication. When a user's domain is configured to use a Microsoft 365 User Verification Source, the user is redirected to the Microsoft 365 authorization URL to complete the login process. Upon a successful Microsoft 365 login, the user's browser is redirected back to SecurityGateway.
Note: SMTP authentication still supports only the “Resource Owner Password Credentials Grant” OAUTH flow, which does not support two-factor authentication.
Example usage:
require ["variables", "securitygateway"];
execute "some-script.bat";
if string "${vnd.mdaemon.execute.exit_code}" "1" {
fileinto "spam";
} elsif string "${vnd.mdaemon.execute.exit_code}" "2" {
reject "This message looks like spam";
}