- The identity of the output, including filename and role.
- Its SHA-256 hash.
- Its size and generation timestamp.
- The case or workflow it belongs to.
- The moment the record was issued.
Verifiable AI records
Verifiable AI Record for reports, JSON files and AI outputs
Generate AI outputs with a verifiable record that proves what was issued, when, and with which documented inputs, without asking anyone to trust HREVN blindly.
Most AI systems generate outputs. Very few allow you to demonstrate how those outputs were produced and whether they have been modified since.
AI outputs are easy to generate, but hard to defend
AI-generated reports, classifications or decisions are often shared without a clear way to demonstrate what system produced them, what data was used, what version of the output is final, or whether the document has been modified since.
In many cases, the problem is not generating outputs, but proving their integrity and origin.
HREVN does not solve that by asking for trust. It solves it by leaving a cryptographic fingerprint that another party can check.
What is a HREVN Verified Record
A HREVN Verified Record is a technical file that certifies the integrity and issuance of AI-generated outputs.
It does not replace the report. It lets you check whether the issued output does or does not match the cryptographic fingerprint recorded at issuance.
What it certifies
What it is not
Its role is specific: to certify the integrity and traceability of the output.
It does not claim by itself that the content is true. What it leaves is a mathematical and cryptographic basis for checking whether the output changed or did not change.
- It does not validate whether the output is legally correct.
- It does not replace legal or compliance review.
- It does not guarantee regulatory compliance.
- It does not contain the full documentation of the case.
How it works
An AI system generates structured outputs such as reports, JSON files or other documents.
HREVN calculates hashes and builds a structured record around those outputs.
A Verified Record is issued as a separate ZIP file.
Anyone can later check whether the output matches or does not match the recorded hash.
The ZIP does not contain the report. It certifies it.
It does not depend on trusting HREVN by authority. It depends on hashes and reproducible cryptographic checks.
Why it matters
In many contexts, the question is not only “What does this report say?” but “Can you prove that this is the exact report that was issued?”
A Verified Record makes that possible.
Typical use cases
AI compliance reports
Internal AI risk assessments
Vendor evaluation outputs
AI-assisted HR decisions
Financial scoring outputs
Documentation delivered to clients
Part of the HREVN workflow
HREVN combines structured AI review, documented outputs and verifiable records to create outputs that are not only generated, but also traceable and defensible.
When the workflow includes multiple documents, photos or supporting artifacts, HREVN can escalate from a lightweight record to a full Evidence Bundle.
Checkpoints and manifests, without remote issuance
The experimental hrevn-workflow SDK lets developers checkpoint AI workflows locally, resume from the last valid step and export a manifest containing hashes.
It is preserved as an independent experiment: it does not issue, sign, anchor or certify HREVN Verified Records or bundles.
Next step
Generate AI outputs with a verifiable record
Use HREVN when you need outputs that can be reviewed, shared and later verified without relying only on trust.