AI governance platform. 648 compliance criteria across four frameworks. Deterministic. Auditable. No AI interprets law.
HEX 165 scans agentic AI systems against 648 compliance criteria across four live modules, with 12 more frameworks in development. One scan produces assessments across every active framework simultaneously.
HEX 165 evaluates your systems against all applicable modules simultaneously. Add new regulations without re-engineering.
12 additional frameworks in development including NIS2, ISO 42001, ISO 27001, PRA SS1/23, HIPAA, SOC 2
No AI interprets law. The compliance engine is static rules derived from legislation. Same input always produces the same output. Auditors can verify every decision.
15 architecture-specific risk flags for multi-agent systems — missing stop mechanisms, dynamic spawning, deceptive instructions, autonomous decisions without oversight.
The HEX 165 agent runs in your environment. Source code never leaves. Only normalised metadata is transmitted. Run offline first, review the output, then decide whether to upload.
Each module is primarily data work, not platform engineering. The shared infrastructure — connectors, rules engine, agents, API — stays the same. Adding a new regulation means structuring its requirements, not rebuilding the platform.