Chainalysis Unveils Ontology Framework to Standardize Blockchain Analytics Evidence

Chainalysis has published a new ‘Ontology’ report that lays out how it classifies blockchain intelligence and what it considers sufficient ‘evidence’ to support those determinations—an effort the company says is designed to make blockchain analytics more verifiable, reproducible, and suitable for high-stakes use in courts, banking compliance, and criminal investigations.

The report, released Tuesday UTC, is positioned as the industry’s first formal disclosure of both a data ontology and an accompanying ‘Evidence’ framework. Chainalysis said blockchain analytics are now routinely used to support judicial outcomes, screen transactions at financial institutions, and help law enforcement trace illicit funds, yet the sector lacks shared standards for objectively testing how a conclusion was reached and whether another analyst could reproduce the same result.

At the center of the issue, Chainalysis argues, is the industry’s long-standing and often inconsistent use of the word ‘cluster.’ For more than a decade, the company said, the term has been applied broadly—sometimes referring to structural identification of address groups and other times to the separate step of linking those groups to a real-world entity. Those are different analytical problems, Chainalysis noted, but they have frequently been treated as the same concept.

That ambiguity matters because the downstream consequences are real. If taxonomy is unclear, two providers can analyze the same blockchain addresses and arrive at different conclusions, potentially affecting transaction monitoring decisions, investigative leads, or even judicial determinations.

Chainalysis’ proposed solution is a tiered model that splits what many firms casually call ‘clustering’ into two distinct layers with different validation expectations. In the company’s ontology, ‘Tier 1’ focuses on structural analysis—identifying which addresses likely belong to the same wallet group. In this stage, the standard is strict reproducibility: another analyst should be able to reach the same grouping, and the process should be auditable by external reviewers.

To preserve deterministic results, Chainalysis said it does not apply machine-learning methods to the foundational step of grouping addresses, arguing that models whose outputs can shift across training runs or parameters are ill-suited for a baseline that should not vary across analysts or over time.

‘Tier 2,’ meanwhile, covers attribution and operator assessment—determining which person, company, or service a wallet group is connected to, and who actually controls it. This is where evidentiary standards become more nuanced. Chainalysis said it references U.S. Intelligence Community guidance, specifically ICD 203, to present not only the basis for a judgment but also the confidence level associated with it—an approach intended to make analytic reasoning transparent rather than purely outcome-driven.

The company emphasized that it is not attempting to impose new rules on the industry or to grade competitors. Instead, it framed the report as a formal documentation of principles it has used internally since its early years—intended to help stakeholders understand “what evidence produced this conclusion” and “whether the same result can be reproduced.”

Chainalysis also pointed to prior external scrutiny of its methods. The company said its analytics passed the ‘Daubert’ standard—used in U.S. federal courts to assess the reliability of scientific evidence—in the case United States v. Sterlingov, and claimed it remains the only blockchain analytics firm to have cleared that threshold to date. Separately, it said researchers at Delft University of Technology involved the company—under its real name—in a verification effort using ground-truth data that was presented at USENIX Security.

In addition to courtroom and academic validation, Chainalysis highlighted operational use by law enforcement, citing a case in which the Seoul Metropolitan Police Agency used Chainalysis Reactor to trace funds during an investigation into an international hacking group. The company presented the example as evidence that methodologies tested abroad are being applied in active investigations in South Korea.

Ultimately, Chainalysis is betting that the next phase of the blockchain analytics market will be defined not only by the sophistication of tracing tools, but by whether their outputs can be defended under scrutiny. As blockchain-derived intelligence influences legal decisions, compliance actions, and criminal probes, the company argued that organizations will increasingly need to evaluate not just the findings but the ‘methodology’ and ‘verification process’ that produced them.

Chainalysis said it hopes the disclosure of its ontology and evidence framework serves as a starting point for improved transparency and trust across the sector.