MEASURING GENERATIVE AI TO GOVERN IT

As generative AI systems scale, governance can no longer rely on declarations alone. It requires public, continuous, and verifiable measurement.

NeoMundi Research is launching an exploratory observatory dedicated to the behavioral study of stability regimes observed across contemporary generative systems.

NeoMundi Observatory

A panel of 10 major generative AI services is observed using a common and reproducible protocol.

The objective is to produce periodic maps of their stability regimes, variations, and behavioral signatures.

All maps are available here.

To receive our Newsletter
(Publications, Mapping, Ratings, Benchmarks, Studies, ...)
Subscribe below.

FREE EXPLORATORY MEASUREMENT

Call the instrument. Measure the drift.

NeoMundi Research can provide exploratory access to its runtime measurement instrument, allowing AI teams, researchers, specialized journalists, or compliance officers to observe the behavior of their own use cases.

One API call. BYOK. Privacy first. No content storage.

You can test stability, drift, validity, hallucination risk, coherence, informational density, and risk profiles on your own prompts, models, agents, or orchestration chains.

The goal is not to sell a promise.
The goal is to measure, document, and discuss the observed signals.

Request an Exploratory Access

OPEN SCIENCE

Researchers, laboratories, PhD students, independent contributors, and AI teams can request Open Science access to the NeoMundi instrument to test, replicate, audit, or challenge its signals on their own use cases.

A measurement instrument must itself be measured, audited, replicated, and challenged.

NeoMundi Research publishes two levels of resources:

Industrial Specifications
Runtime architecture, signals, interoperability contracts, OBS/GOV, BYOK, privacy, governance artifacts.
→ github.com/neomundi-io

Scientific Program
Datasets, protocols, scripts, scoring methods, anonymized results, and interpretation limits.
→ Available 06/06/2026

ORIENTATION COMMITTEE IN FORMATION

NeoMundi Research is progressively forming an orientation committee to challenge the method, measurement protocols, and interpretation limits of the instrument. The first members bring an external perspective on the scientific, technical, ethical, and institutional dimensions of the project. The committee is intended to expand with complementary profiles: researchers, AI specialists, lawyers, compliance experts, public actors, and audit professionals.

First Members:

  • Joël Ignasse, Scientific journalist for Sciences et Avenir and La Recherche. Author of “Sur les traces du Nouveau T.rex” (Eyrolles, 2025).
  • Cédric Chatelain, Quality consultant and IRCA-certified auditor, specialized in audit, validation, and ISO standards. Based in Bern, Switzerland.
  • James Aull, Founder of ASRO™ – AI Systems Reliability Operator, focused on governed-state witness evidence, AI governance, and attestation frameworks. Based in Twin Lake, Michigan, United States.

Scroll to Top