Scientific Contributor / Data Scientist – NeoMundi AI Observatory
Most discussions on AI governance start too late.
We believe measurement must come first.
We have developed ControlTower, a diagnostic, piloting, and traceability instrument designed to continuously observe the risk of AI responses in real-world conditions.
ControlTower enables standardized and reproducible measurement campaigns, allows comparison of behaviors between models, and tracks their evolution over time. It makes visible critical phenomena that current benchmarks struggle to detect reliably: response stability, variability, potential drifts, risk signals, traceability, and behavioral regimes.
We are now launching the NeoMundi AI Observatory
An independent program dedicated to measuring, documenting, and understanding the behaviors of generative AI models.
Objective
To progressively build an open, rigorous, and useful framework to better measure AI systems before claiming to govern them.
As part of the Exploratory Cycle 2026, we are looking for a limited number of scientific contributors, data scientists, AI researchers, AI engineers, compliance, audit, risk management, and algorithmic governance profiles.
Contributions may include
- Analysis of measurement campaigns;
- Methodological review;
- Protocol design;
- Documentation of limitations;
- Interpretation of observed signals;
- Linking results to compliance requirements, particularly the EU AI Act.
Contributors will have the opportunity to participate in building an emerging AI measurement framework, access structured campaigns and datasets, contribute to aggregated publications, and help define methods that could influence future standards of traceability, audit, risk, and governance of AI systems.
Initial commitment can remain light: protocol review, ad-hoc analysis, methodological contribution, scientific critique, or participation in a measurement campaign.
The Observatory will produce regular publications: cartographies (AI mappings), barometers, methodological notes, public references, and aggregated analyses by provider, model, and industry sector.
→ Browse Observatory Publications
The Observatory follows a privacy-first approach, with aggregated publications, scientific caution, traceable methods, and clear documentation of limitations.
The Observatory is supported by Infomaniak, a leading European player in ethical cloud and sovereign infrastructure.
To contribute or learn more: https://neomundi.org/en/submit-a-contribution