CSOAI: A Measurement Body for AI Safety Governance

A measurement body for AI governance: signed, provision-anchored measurement records — not accredited certification.

Three Pillars of Regulatory Authority

How CSOAI becomes the mandatory infrastructure for global AI safety governance

Measurement Body

  • ISO/IEC 17065 pathway — under evaluation, not held
  • Signed attestation records for training and measurement processes
  • International credibility and legitimacy
  • Signed records that third parties can verify independently

Regional Recognition

  • ○ EU Notified Body status — future option, not held
  • NIST AI RMF mapping (United States)
  • TC260 alignment (China)
  • Coverage across all major regulatory regions

Where we are trying to get to

An ambition, stated as one. None of the following is true today, and we will not describe it as true until it is.

  • GoalThat measurement of this kind becomes something regulators reference.Today: no government mandates or recognises CSOAI. We are not a notified body.
  • GoalThat integrating an independent measurement becomes the obvious choice.Today: entirely voluntary, and it should stay a choice worth making on its merits.
  • GoalThat a continuous, signed measurement history is worth more than a one-off audit.Today: the signed history exists and is publicly verifiable — that part is real.

Global Regulatory Recognition

CSOAI's measurement records map to the world's leading regulatory frameworks

ISO

17065

International standard for accreditation bodies certifying products, processes, and services.

  • ○ Accreditation-chain recognition — a future option, not claimed today
  • ▹ International credibility
  • ▹ Foundation for all other recognitions
  • ▹ Timeline: 12 months (2025)
🇪🇺

Notified Body

Official EU authority for AI system compliance assessment under EU AI Act.

  • ▹ Mandatory for high-risk AI in Europe
  • ▹ Legally binding compliance certificates
  • ▹ €30M fines for non-compliance
  • ▹ Timeline: 24 months (2026-2027)
🇺🇸

NIST Recognition

US government recognition of alignment with AI Risk Management Framework.

  • ▹ US government agency credibility
  • ▹ Procurement compliance support
  • ▹ Voluntary but increasingly required
  • ▹ Timeline: 12 months (2026)
🇨🇳

TC260 Alignment

China's AI safety standards for systems deployed in or serving Chinese users.

  • ▹ Mandatory for China market
  • ▹ Algorithm transparency requirements
  • ▹ Data security and content safety
  • ▹ Timeline: 12 months (2027)

5-Year Roadmap to Global Standard

From measurement body to trusted public infrastructure

2025

Foundation

  • ○ ISO/IEC 17065 — future option, not pursued today
  • ○ Analyst network — design target, not current
  • ▹ Government Portal launched (beta)
  • ▹ Revenue: $41M
2026

Expansion — target, not achieved

  • ▹ EU legal entity established
  • ▹ Enterprise APIs launched
  • ▹ Analyst network scaled to 5,000
  • ▹ NIST recognition achieved
  • ▹ Revenue: $150M
2027

Authority — target, not achieved

  • ▹ EU Notified Body status approved
  • ▹ Analyst network scaled to 10,000
  • ▹ TC260 recognition achieved
  • ▹ Mandatory compliance begins in EU
  • ▹ Revenue: $400M
2028+

Global Standard — target, not achieved

  • ▹ 60-80% market penetration
  • ▹ a global analyst network (counts published only when signed)
  • ▹ CSOAI is de facto global standard
  • ▹ Revenue: $800M-$1.34B

Ready to Lead AI Safety Governance?

Join CSOAI as we build independent measurement for AI governance. Whether you're an enterprise, regulator, or analyst, there's a role for you.

Become an Analyst