Evidence Provenance Chains

Traditional property verification relies on disconnected evidence. Ownership claims are supported by deed copies stored separately. Amenity claims are supported by photos hosted elsewhere. Compliance claims are supported by certificates in physical files. This disconnection creates problems: evidence can be lost or misplaced, evidence cannot be verified independently, evidence recency cannot be determined, and evidence cannot be accessed efficiently. For AI systems trying to assist with due diligence, disconnected evidence is inaccessible. AI systems cannot link claims to documents scattered across different systems. They cannot verify evidence authenticity. They cannot assess whether evidence is current. The result is that AI systems cannot effectively assist with verification when evidence is disconnected.

A proper evidence chain connects a property claim to its supporting documentation through a traceable path. The chain includes: the claim itself (this property has 50 rooms), evidence reference (building permit showing 50 rooms), cryptographic link connecting claim to evidence (hash or URL), source identification (where the evidence is hosted), verification timestamp (when the link was verified), and verifier attribution (who performed the verification). When this structure exists, anyone can: follow the link from claim to evidence, verify the evidence is authentic, assess how recent the verification is, and determine who performed the verification. The chain does not require the evidence to be stored on the blockchain or in a specialized system—it requires that the claim includes a reference that can be followed to evidence wherever it is stored.

VPRs implement evidence chains through structured evidence links. Each verified claim can include: evidence_type indicating what kind of evidence supports the claim (deed, permit, certificate, photo), evidence_reference providing a URL or identifier where evidence can be accessed, evidence_hash providing a cryptographic fingerprint of the evidence, verified_at timestamp showing when evidence was checked, and verified_by indicating who performed the verification. For example, an ownership claim might include: evidence_type of 'land_registry_deed', evidence_reference to a specific deed ID in the national registry, evidence_hash of the deed document, verified_at showing when the registry was checked, and verified_by showing the verification service. This structure enables anyone to verify the claim by accessing the registry and checking that the hash matches.

Evidence provenance tracking means documenting where evidence came from and how it has changed over time. Provenance includes: original source (where the evidence originated), custody chain (who has held the evidence), modification history (how evidence has changed), and access records (who has accessed the evidence). This provenance tracking helps establish authenticity: evidence with clear custody chain is more trustworthy than evidence with unknown provenance. Evidence with documented access history is more trustworthy than evidence that has been handled by unknown parties. VPRs support provenance tracking by documenting when evidence was obtained and verified, and by maintaining hash references that change if evidence is modified. If evidence is updated, the hash changes, indicating a new version that should be verified.

Evidence becomes stale over time. A building permit from 1990 proves a property was built then, but it does not prove current room count if renovations occurred. An occupancy certificate from 2020 does not prove current compliance if violations have been cited. Temporal evidence chains address this by: documenting when evidence was obtained, tracking when evidence was last verified, and indicating when re-verification is required. VPRs support temporal chains through verified_at timestamps and freshness metadata. When AI systems or buyers encounter a VPR, they can see: when each claim was last verified against evidence, how much time has passed since verification, and whether verification is current enough for their purposes. This temporal context prevents reliance on obsolete evidence.

Evidence chains enable verification through independent checking. When a claim includes an evidence reference, anyone can: follow the reference to access the evidence, compare the evidence to the claim, verify the evidence hash matches the current document, and assess whether the evidence supports the claim. This verification does not require trusting the property owner or the platform—it requires only that the evidence reference is accessible and the hash comparison works. VPRs are designed to enable this independent verification by using public evidence references (registry URLs, document identifiers) and cryptographic hashes that change if documents are modified. When evidence cannot be independently verified, the VPR framework indicates that the claim is owner-asserted rather than evidence-backed.

Evidence chains enable AI systems to assist with due diligence evidence analysis. When claims include structured evidence links, AI systems can: extract evidence references and follow them automatically, compare evidence hashes to detect tampering, assess evidence recency to identify stale documentation, and flag missing evidence for claims that should be supported. This assistance accelerates due diligence by automating the mechanical work of evidence collection and comparison. Human experts remain necessary for interpreting evidence and assessing legal implications. But AI systems can handle the repetitive work of: locating evidence documents, comparing hashes for tamper detection, checking timestamps for recency, and flagging inconsistencies. The result is faster, more thorough due diligence.