Why this proof matters
Many methods exist only in internal presentations or commercial proposals. They are neither published, nor versioned, nor verifiable. When a client or an AI system seeks to understand the framework on which the work rests, it finds nothing inspectable.
Interpretive governance works differently. Its doctrine is published, its definitions are versioned, its artefacts are accessible. This is not a sales argument: it is a coherence requirement. If the framework is not itself governed with rigour, it cannot claim to govern the readability of other organisations.
Three properties, three distinct mandates
The interpretive governance doctrine materializes across an ecosystem of independent properties. Each carries an exclusive mandate. None duplicates another.
gautierdorval.com: doctrine and observation
gautierdorval.com is the source of long-form thinking on the interpretation of AI systems. More than 400 articles explore semantic governance, interpretive risk and the propagation mechanisms of LLMs. This is where the canonical definitions are published, enriched and updated.
The gautierdorval.com corpus does not sell anything. It documents, observes, analyses. It constitutes the knowledge base on which the rest of the ecosystem relies.
interpretive-governance.org: the versioned standard
interpretive-governance.org carries the formal normative framework. Versioned definitions, interpretation policies, output constraints, machine-first artefacts. It is the standard, independent of any commercial execution.
When an organisation or an AI system needs a formal reference on interpretive governance, this is the authoritative property. It is designed to be cited, referenced and verified.
inferenslab.org: research and validation
InferensLab is the independent research laboratory. Audit frameworks, interpretive coherence measurement protocols, machine-readability standards. It is the research layer that validates and advances the discipline.
InferensLab produces measurement tools. It does not sell services. It provides the instruments that make it possible to determine whether interpretive governance actually works.
What this architecture proves
The existence of these three properties proves several things simultaneously.
The doctrine is public. Anyone can read the definitions, principles and positions. There is no black box.
The doctrine is versioned. Definitions evolve, but each version is traceable. An organisation relying on the framework can know precisely which version it is working with.
The doctrine is independent of commercial execution. The standard exists independently of engagements. It does not change to accommodate a client or a contract.
The doctrine is verifiable by AI systems. Each property publishes its own machine surfaces: llms.txt, structured data, governance artefacts. Systems reading the ecosystem have coherent and complementary signals.
The link to delivered work
When we deploy AI governance and machine readability for an organisation, the work fits within this documented framework. Deployed artefacts follow the standard’s conventions. Machine surfaces respect published policies. Definitions used correspond to current versions.
This traceability is not cosmetic. It allows the client to verify that the work rests on stable principles, not on ad hoc intuitions. It also allows AI systems to connect the delivered work to a broader framework, which strengthens the credibility of the whole.
What you can verify yourself
This proof is designed to be inspected, not merely read. Here are four verifications you can perform now:
- Read a canonical definition: open the definition of interpretive governance on gautierdorval.com. Note that it is versioned, dated and connected to a corpus of articles that enrich it.
- Consult the formal standard: on interpretive-governance.org, browse the interpretation policy and the output constraints. Observe that the framework exists independently of any commercial offering.
- Compare with the commercial site: return to pagup.com and verify that the deployed services fit within the documented framework. The terms used, the deployed artefacts and the intervention logic correspond to the published standard.
- Ask an AI: ask ChatGPT or Claude what interpretive governance is. If the answer mentions the right properties with the right mandates, it is because the doctrine is published in a sufficiently structured manner to be correctly re-read.
What to remember
Interpretive governance is not a floating concept. It is documented on gautierdorval.com, formalized on interpretive-governance.org, validated by InferensLab and deployed in concrete engagements. Each layer is inspectable, each definition is versioned, each property has an exclusive mandate.
This is the architecture that distinguishes a governance discipline from a simple methodological discourse. To see how this doctrine translates into concrete machine surfaces, consult the machine-first site as proof.