− Our edge
Our Manifesto
− Illustrative dependency mapn
See what a source revision puts back in review.
Switch source versions and select a node to see how the claim and its tracked uses are affected.
Version 1 is the source for the 120 °C claim.
− Original Manifesto
A manifesto for
Governed Claims Infrastructure.
01. Claims are written, reviewed, corrected, and rewritten every day.
A scientific statement is approved in one document, adapted for another market, shortened for another channel, and reused months later in a different communication.
The evidence may remain unchanged. The wording may change. The context may change. The regulatory environment may change.
Yet the claim itself is rarely treated as a governed object.
It should be.
02. Evidence should not disappear into documents
Scientific and regulatory evidence is the foundation of regulated communication. But once that evidence is incorporated into documents, presentations, promotional materials and other content, its relationship with the claims derived from it becomes increasingly difficult to maintain.
Teams search for the source of a number.
They reconstruct why a statement was approved.
They discover that the same claim exists in several versions.
They review content again because they cannot be certain which underlying evidence still applies.
This is not primarily a people problem.
It is an infrastructure problem.
The industry has built sophisticated systems to manage documents, content and review workflows. But the scientific Claim that connects evidence to communication remains fragmented across them.
03. The Claim is the missing layer
A Claim is more than a sentence.
It is an assertion grounded in evidence, used in a defined context, subject to specific constraints and potentially reused across markets, channels and communications.
Once understood this way, the Claim becomes something that can be:
constructed, traced, governed, versioned, reused and continuously monitored.
This is a different operational model for regulated communication.
Instead of asking only whether a document is approved, organizations can govern the Claims from which their communications are built.
04. AI makes this shift necessary
Generative AI is changing the economics of content creation.
It can produce more content, across more channels, in more markets, at a speed that traditional review processes were never designed to accommodate.
But AI does not solve the underlying governance problem.
A language model is probabilistic. It can generate a plausible statement that is unsupported by the evidence.
The answer is not to expect the model to become the authority on what an organization is allowed to say.
The model can generate. The governance layer must decide what can be released.
This requires a clear separation between generation and governance: authoritative evidence and governed Claims provide the boundaries within which AI can operate.
05. Compliance must move upstream
Traditional compliance processes often identify problems after content has been created.
A different model is possible.
Scientific evidence can be structured into traceable Atoms. Those Atoms can provide the building blocks from which Claims are constructed. Constraints can govern how those building blocks may be combined. Claims can then be independently verified before they are released.
Compliance is no longer simply a downstream review step.
It becomes part of the architecture.
This is the principle of Compliance by Design.
06. Claims should remain connected to their evidence
Governance cannot end when a Claim is approved.
Scientific evidence changes. Regulations evolve. Labels are updated. New studies are published. A Claim that was valid yesterday may require review tomorrow.
A governed Claim therefore needs a continuing relationship with its underlying evidence and with the communications that use it.
When evidence changes, the Claims affected should be identifiable.
When a Claim changes, the communications that depend on it should be visible.
When a Claim is no longer valid, its use should not continue unnoticed.
This is Continuous Claim Governance.
07. A new infrastructure for regulated communication
The industry has already created infrastructure around other fundamental enterprise objects.
- ERP systems govern transactions.
- CRM systems govern relationships.
- Content and MLR systems govern documents and review processes.
The next question is whether regulated communication also needs an infrastructure that governs the Claim itself.
We believe it does.
Governed Claims Infrastructure provides the missing layer between scientific evidence, Claims and the communications built from them.
It does not replace existing enterprise systems.
It connects them through a governed object that can carry evidence, context, status and accountability throughout its lifecycle.
08. Starting with Life Sciences
Pharmaceutical and life sciences organizations are where this need is most immediate.
The combination of complex scientific evidence, stringent regulatory requirements, multiple markets, omnichannel communication and rapidly increasing content volumes makes document-centric governance increasingly difficult to scale.
But the underlying principle extends beyond Life Sciences.
Wherever organizations communicate from authoritative evidence and where unsupported claims create material regulatory, financial, legal or safety risk, the Claim can become a fundamental unit of governance.
Pharma is where the transformation begins.
09. The shift
The future of regulated communication is not more documents.
It is not more review cycles.
And it is not asking AI to become the source of truth.
It is building an infrastructure in which evidence remains connected to the Claims it supports, Claims remain governed throughout their lifecycle, and AI operates within clearly defined boundaries.
10. Documents have approval cycles. Claims need ongoing governance.
That is the shift toward Governed Claims Infrastructure.
− COMPLIANCE BY DESIGN
Regulated work will move at machine speed.
Governance must too.
Validix turns approved evidence, policy and context into an executable control plane for generated language.
So regulated enterprises can create at AI speed — without trusting AI to decide what is valid.
− Continue exploring
About Us →
Explore how the same approach applies across industries.
Insights →
Read more about evidence and AI governance.