Abnormal AI for Government
Abnormal AI for Government delivers AI-native human behavior security to federal and regulated public sector environments, protecting against socially engineered threats across email and connected cloud applications.
This resource defines the security and compliance controls governing Abnormal’s FedRAMP-authorized environment.
Security and Compliance Overview
Dedicated Authorization Boundary
All model processing and data handling remain confined within the FedRAMP authorization boundary in a dedicated Government tenant hosted exclusively in AWS GovCloud.
Phishing-Resistant Administrative Access
Hardware-backed authentication with FIPS-compliant YubiKeys and integration with SAML SSO and PIV/CAC strengthens privileged access controls.
Per-Tenant Behavioral Isolation
Behavioral baselines are enforced individually by tenant to maintain strict data separation while enabling high-fidelity anomaly detection.
Defined Operational SLAs
Documented vulnerability remediation timelines and formal incident response testing ensure predictable, auditable security operations.
The complete Abnormal AI Security and Compliance Overview is available below.
