Evidence integrity
What is evidence integrity: evidence integrity explained
Evidence integrity is the set of technical and procedural controls that let an organisation prove exactly what digital material it received, when the receiving system recorded it, whether that material is still unchanged, and what the team did with the case afterwards.
What does evidence integrity mean in practice?
It means a claim about custody and handling, not about truth. Evidence integrity describes how a file travelled and whether it stayed intact. It says nothing about whether the dent in the photo happened the way the customer described it.
A handler reopening a disputed file has a narrow set of questions, and every one of them is about the record rather than the incident:
- What did we receive? The original video, photos, answers and spoken explanation, not a compressed copy someone re-saved and forwarded.
- When did we receive it? A time recorded by the receiving system, not read off the sender's phone clock.
- Is this still the same file? A cryptographic check that the content has not changed since arrival.
- What happened during capture? A timeline of session events, from the link opening to the submission completing.
- What happened after it arrived? Who reviewed it, who it was assigned to, which status changes and retake requests followed.
An email attachment answers almost none of that. It tells you someone supplied a file.
How does evidence integrity work?
Integrity is not one feature. It is a stack, and each layer closes a gap the previous layer leaves open.
- Capture instead of upload. Material recorded inside the requested flow, rather than picked out of a gallery, shortens the path between what happened and what the organisation receives.
- Server-side receipt time. The receiving system stamps arrival, so the record does not rest on a device clock anyone can change.
- A digital fingerprint. A SHA-256 hash of each file on arrival, so any later alteration becomes detectable.
- A session timeline. Recorded events across the capture session, in server time.
- Control points. Technical signals such as device history, IP reuse, virtual camera detection and recording location, graded for human attention rather than acted on automatically.
- A signed seal. A cryptographically signed manifest around the whole submission, checkable outside the platform that produced it.
- An audit trail. The case history after arrival: received, reviewed, assigned, status changed, retake requested, closed.
Evidence integrity example: the claim reopened eight months later
A motor claim is settled in March. In November the insured disputes the repair scope and a solicitor asks which images the original decision rested on. The integrity record can show the four files received on 3 March at 14:12 server time, confirm their SHA-256 fingerprints still match, and list who reviewed them and when a retake was requested.
What it cannot show is whether the scuffed bumper in file two was damaged in that incident or a fortnight earlier. That question belongs to the assessor, not to the record.
What evidence integrity does not prove
This is where a lot of vendor copy overreaches, so be exact about the ceiling.
- It does not prove what the file depicts. A verified, unaltered recording can still show a staged scene, an unrelated object, or damage that predates the policy.
- It does not prove when the event happened. A server-side receipt timestamp proves when the submission reached the system. That is all it proves.
- It does not prove who caused the damage, or why, or whether the account given around the capture is accurate.
- It does not establish intent. Technical signals raise questions. People answer them.
Harder to manipulate is not the same as impossible to deceive. A policy document that blurs those two sets a claims team up for an argument it cannot win in front of an ombudsman.
When does evidence integrity matter?
Fabricated imagery stopped being a specialist problem. Aviva reported detecting more than 18,400 suspect claims worth 233 million pounds in 2025, singling out AI-generated accident photographs and manipulated documents as a growing pattern in motor claims, with the value of detected motor fraud up 39% year on year (Aviva, June 2026).
The background sits in our explainer on AI-generated insurance fraud and in the breakdown of what makes photo evidence in insurance claims hold up or fall apart. The workable response is not to distrust every customer. It is to collect evidence that carries its own provenance, which is the argument behind structured insurance claim documentation.
Treat it as a policy question before a product question. Decide what your team must be able to reconstruct, how long you need to keep it, who may verify it, and what happens when a signal fires. Data protection obligations pull against indefinite retention, so retention windows belong in the same conversation as the integrity design.
Venta Capture, a product of VentaVid, builds this layer into every submission: 28 control points per submission, made up of 10 automatic signals, 21 recorded session events and 1 signed seal, plus SHA-256 fingerprints on arrival, a signed seal that can be verified with a public key, GPS location, and GDPR-compliant storage with configurable retention.
