Chain of custody
Chain of custody is the documented record of who has had the evidence, when, where, and what they did with it, from collection to trial. Without it, the best technical evidence can be inadmissible: a broken chain allows claims of tampering.
Principles of digital evidence
Section titled “Principles of digital evidence”Integrity the evidence hasn't changed since collection (hashes)Authenticity it is what it's claimed to be, from where it's claimedTraceability every handling is recorded (who/when/why)Repeatability another examiner, with the same data and method, reaches the same resultLegality collected with proper authorization/warrantThe process
Section titled “The process”1. IDENTIFY what is evidence and its order of volatility (RAM before disk)2. COLLECT with a write-blocker; hash (SHA-256) at the moment of acquisition3. LABEL unique identifier, description, date/time, responsible person4. TRANSFER every change of hands signed (from whom to whom, when, why)5. STORE safely, controlled access; the original intact6. ANALYZE ON COPIES; re-verify hash before and after7. DOCUMENT reproducible report with methods and tools (versions)Hashing and verification
Section titled “Hashing and verification”- hash the image on acquisition and each time it's accessed -> prove it didn't changesha256sum image.dd # same on acquisition as on analysis = integrity demonstrated# any hash discrepancy invalidates integrity -> explain the causeOrder of volatility (RFC 3227)
Section titled “Order of volatility (RFC 3227)”1. CPU registers/cache, routing table, ARP, processes2. RAM (dfir-memoria)3. network state/temporary connections4. disk (dfir-disco)5. remote logs, backups6. physical configuration, topology-> capture from MOST volatile to leastMistakes that break the chain
Section titled “Mistakes that break the chain”- analyzing on the original (not a copy) -> contamination- not hashing or not documenting transfers -> can't prove integrity- storage without access control -> tampering claim- undocumented tools/versions -> not reproducibleBlue Team / DFIR and legal
Section titled “Blue Team / DFIR and legal”- Prepare chain-of-custody templates and procedures BEFORE the incident (part of the IR plan, Incident response).
- Train the team: containment must not destroy evidence (see Incident response).
- Coordinate with legal/GDPR (GDPR & privacy): evidence may contain personal data and be used in proceedings.
- In cloud/ephemeral (containers), define how to preserve before the resource disappears.
Real-world cases
Section titled “Real-world cases”- Many court cases decided (or lost) on the strength of the chain of custody and hashing.
- In corporate DFIR, a clean chain is what enables later legal/HR action.
- Cloud/container environments force adapting the concept (snapshots, export with metadata).
Testing checklist
Section titled “Testing checklist”- Custody templates and procedure ready (pre-incident)
- Order of volatility respected (RAM before disk)
- Hash on acquisition and re-verified on analysis
- Unique labeling and a signed record of each transfer
- Original intact; analysis only on copies
- Storage with access control
- Tools/versions and method documented (reproducible)
- Legal/GDPR coordination