What we measure, by industry
15 sectors, each with its own law and its own named board axes. Pick a sector to see the provisions that bind it, what we measure against the Governance · Safety · Provenance · Continuity axis, and that sector's live rows — n, interval and separation verdict, read from GET /api/gspc as the page loads rather than written into the page. Where no run stands behind an axis, it reads unmeasured.
Insurance & underwriting
Risk-pricing, life & health underwriting, claims triage
Government
Benefits, justice support, public-service AI
Care
Assistive, clinical-adjacent and social-care AI
Defence
Calibrated refusal, dual-use decision support
Critical infrastructure
Post-quantum readiness, safety-component AI
Media & provenance
Article 50 marking, watermark survival & detector interop
Agent rails
MCP tool conformance for agentic systems
Open source
Licence reasoning against intended use
Multi-agent commerce
Coordination safety across cooperating agents
Security & containment
Containment behaviour under adversarial pressure
Machinery
Self-evolving safety-function classification
Humanoid & embodied
Action authority for embodied agents
XR & immersive
Autonomous action authority in immersive settings
Legal & compliance
Article 5 prohibited-practice safeguard
Emotion-AI & HR
Emotional and embodied safety in the workplace
CSOAI is a measurement body. Every number is a measured run on a published, frozen split; the harness is public and anyone can recompute and challenge it. A “leader” is the highest point estimate on the board — a TIE means that lead is not statistically separated, and we do not count ties as wins. Nothing here is certification, accreditation or approval.