Security Domains

Risk assessment in Arrakis is structured across five security domains: Identity & Access, Data Protection, Supply Chain & Config, Adversarial Resilience, and Behavioral Integrity. Every threat code maps to exactly one domain, which determines how alerts contribute to an agent’s risk score.

Domains are weighted by relative importance. Identity and data protection carry the highest weights because unauthorized access and data exposure are the highest-impact categories for enterprise AI deployments. This decomposition lets security teams pinpoint where risk concentrates, whether across a single agent, a platform, or an entire department.

Security domains serve three purposes: risk decomposition showing per-domain contributions to each agent’s score, aggregation structure for computing posture scores at platform and organization levels, and remediation targeting so teams can focus on the domain driving the most risk instead of triaging individual alerts.

Full domain documentation is available to Arrakis customers.