Digital Asset Claims Evidence Verification
Evidence verification is the gate every finding passes through before it is written into a case record. It exists to separate a single-source observation from a corroborated fact, and to document the difference explicitly rather than presenting both with equal weight.
What is checked
Source provenance
Every piece of source material — a registry filing, a block explorer record, a captured page — is checked for its origin, retrieval date and whether it can be independently re-obtained by a reviewer, not just cited from memory or a summary.
Independent-source corroboration
A finding drawn from one source class — for example, a single OSINT match — is tested against a second, independent source class such as on-chain data or infrastructure records before it is upgraded from a lead to a supported finding.
Chronology and consistency checks
Dates, sequences and stated facts are checked against each other for internal consistency; a transaction that could not physically have occurred before its cited cause is flagged and re-examined rather than left in the record.
Grading
Verified findings are graded from verified through strongly supported, partially supported, unverified, conflicting, to insufficient evidence, and every case report states which grade applies to each material finding rather than presenting a single undifferentiated narrative.
Limitations
AI outputs are treated as investigative leads until corroborated — a machine-generated flag or cluster is never recorded as verified until an analyst confirms it against independent source material. Verification reduces error; it cannot manufacture evidence that does not exist, and questions the available records cannot answer are recorded as unresolved rather than guessed at.