One Health increasingly depends upon connection. Animal-health data needs to connect with environmental information, livestock movements with disease surveillance, and food production with traceability, while laboratory results, climate information, supply chains, research data and public-health systems all contribute pieces of the same picture.
The Food and Agriculture Organization of the United Nations took a significant step in this direction in September 2026, launching its Sustainable Livestock Transformation and One Health Intelligence & Knowledge Hub at the Global Conference for Actions on One Health in Agrifood Systems in Rome.1 The Hub brings together livestock data, analytics, tools and knowledge, organised through dashboards by theme and geography covering production, animal-health risks, antimicrobial resistance, climate interactions and socioeconomic dimensions, with a Livestock AI Assistant to help users navigate it. FAO describes the objective as helping turn data into actionable intelligence and support better decision-making.
It is precisely the direction in which One Health needs to move. But as information becomes more connected, another problem becomes more important: how do we know that the evidence we are connecting can be trusted?
Combining datasets can reveal patterns that would remain invisible within individual systems. An animal-health event becomes more meaningful alongside livestock movement records, environmental surveillance becomes more useful when linked to meteorological information, and a laboratory finding matters more when connected to a particular animal, holding, shipment or location.
That is the promise of integrated intelligence, but integration also creates dependency. A conclusion produced from five datasets depends upon the quality and integrity of all five, so if one record is incorrect, incomplete, misattributed or altered without detection, the resulting analysis may be wrong in ways nobody can see.
Consider an animal-health record. It may tell us that an animal was tested for a particular pathogen, but for that information to become reliable evidence we also need to know who collected the sample and when, which animal it originated from and how that animal was identified, where it was located, which laboratory performed the test and by which method, whether the result was amended, who entered the record, and whether anything changed after creation. That is the difference between having a record and understanding the provenance of that record.
A provenance system cannot guarantee that a biological observation is correct. If somebody records an animal’s identification incorrectly, technology cannot determine which animal they intended to identify, and if a sample is contaminated before it enters a laboratory information system, no digital ledger can repair the sample.
What provenance can do is preserve evidence about where information originated, who supplied it, how it has changed and how it relates to other records. That creates accountability and makes errors easier to investigate, which is a narrower claim than the one usually made for this kind of technology, and a more defensible one. We set the distinction out in full in our frequently asked questions, because it is the point most often blurred.
An outbreak investigation may need to combine animal identification records, farm records, veterinary observations, movement records, laboratory samples, diagnostic results, genomic data, feed information, environmental samples, slaughterhouse records and information from neighbouring holdings, each originating in a different system maintained by a different organisation. If the relationships between those records have to be reconstructed manually during an emergency, valuable time is lost. A stronger architecture establishes those relationships before the emergency occurs.
Reliable provenance requires persistent identifiers capable of connecting records to the entities and events they describe. Without reliable identity, interoperability alone is not enough: two databases may technically exchange information while remaining unable to establish confidently that they are describing the same animal, sample or event.
AI can dramatically improve our ability to identify relationships across large and complex bodies of evidence, and FAO’s inclusion of an AI assistant in the new Hub reflects where the field is heading. But AI does not remove the need for trustworthy source data; it increases it. When an automated system draws conclusions from millions of interconnected records, the provenance of those inputs becomes critical, and we need to be able to distinguish between what the analysis says and what evidence the analysis is based upon. One should always be traceable back to the other.
Integrated One Health intelligence does not require every organisation to place all of its information into one enormous central database. Different organisations have legitimate reasons to retain control over their own information, including commercial confidentiality, research governance, personal and farm data protection, and national legal requirements. The challenge is therefore not to centralise data but to make evidence interoperable while maintaining appropriate governance and control.
At The BioChain, this is the problem we are addressing. The BioChain is designed as a verifiable provenance layer connecting records across otherwise separate systems, and the aim is not to replace laboratory systems, farm-management software, government databases, research repositories or supply-chain platforms. Our model runs in five steps, set out in full on our technology page.
The five steps are the same everywhere. Here is what each one connects in this context.
The BioChain does not declare that every record is factually correct. It makes the origin and history of that record verifiable, which is a different and more honest promise. The Ancient BioChain runs all five steps in the open on public research data, where the chain and its gaps can be inspected by anyone.
The same architecture can contribute to animal-disease surveillance, outbreak investigation, food provenance, livestock traceability, biobank management, research reproducibility, environmental monitoring, pharmaceutical and life-science evidence, and cross-organisational One Health collaboration. In each case the common problem is not simply storing information, but maintaining trust as information moves between organisations and purposes.
For many years One Health has correctly emphasised that human, animal and environmental health are interconnected, and the next stage is operational. If the systems are interconnected then the information describing those systems must also become interoperable, and once information becomes interoperable, provenance, identity, authentication and governance become central questions rather than technical footnotes.
Our colleagues at One Health Security have approached the same FAO launch from the other direction, asking what a One Health early-warning system should actually be watching across nine interacting domains. The two questions belong together: knowing what to watch is of limited use if the evidence arriving from each domain cannot be traced back to where it came from.
We should be able to establish where a record originated, when it was created, who or what created it, what entity it refers to, whether it has changed, how it relates to other records, and whether the evidence presented now is the same evidence that was originally recorded. During routine operations this improves accountability, during an outbreak it saves time, and during a dispute it helps establish what actually happened. When AI or automated systems are making increasingly complex connections, it provides a route back from the conclusion to the underlying evidence.
The future of One Health intelligence will involve more data, better analytics, artificial intelligence and greater connection between animal health, human health, agriculture, environment and supply chains. But connection alone should not be the destination. The objective should be connected evidence: evidence whose origin can be established, whose history can be examined, that can cross organisational boundaries without losing its provenance, and that decision-makers can interrogate when the consequences matter.
Because in an emergency, knowing something is valuable. Knowing why you can trust it may matter just as much.
For the same argument applied to a concrete programme, see Digital Livestock Traceability Is Improving, which looks at what the UK’s investment in livestock identification does and does not solve.
No, and it should never be sold as though it does. It can establish who issued a record, what it relates to, which method produced it and whether it has changed since. Whether the underlying test was performed correctly rests on accreditation, quality assurance and inspection, which are different controls entirely.
No. Organisations have legitimate reasons to keep control of their own information, including commercial confidentiality, research governance and data protection. The BioChain connects existing systems and records the relationships between their records; those systems remain the authoritative source of their own data.
An audit trail records what happened inside one application. A provenance layer records what happened between applications and between organisations. If data and its whole history stay within a single trusted system, a good database and audit trail may be all that is needed; the problem appears at the boundary.
Key takeaways
If your organisation generates, manages, analyses or governs biological evidence and would be interested in participating in a UK or European provenance demonstrator, The BioChain would welcome the conversation.
Get in touch
Regulatory developments, technical notes and platform news — sent occasionally, straight to your inbox.