News · August 20, 2026

The Orcher immutable audit ledger enters external review

Evidence that only its author can verify is not evidence. The ledger is being tested by people who did not build it.

By Dipp AI Research — The editorial desk behind the Dipp AI record

Dipp AI has placed the Orcher immutable audit ledger into external review: its tamper-evidence construction, per-action record schema, retention semantics and export format are being examined against the questions auditors and regulators actually ask.

The Immutable Audit Ledger is the component of Orcher that makes every other component provable. It has now entered external review. The reason is straightforward: a record whose integrity can only be attested by the party that produced it is a claim, not evidence, and an accountability product that expects to be believed on its own word has misunderstood its own market.

What is being reviewed

Tamper-evidence construction — whether an entry can be altered, reordered or removed without detection, and how detection surfaces.
Record completeness — whether the per-action schema captures the directive, authorising role, routing decision, boundary decision, realised cost and outcome without gaps.
Write semantics — whether an action can complete when its ledger write fails, and what the system does when it cannot.
Retention and legal hold — whether records survive the operational pressures that usually erode logs: cost policy, migration, and re-platforming.
Export — whether an auditor receives something they can independently test, rather than a report they must take on trust.

FIG. A03-01

The Verified Execution Cycle

Tap a stage

Directive

A named professional states intent. Nothing runs anonymously.

Fig. 1 — The verified execution record. Directive, authority, routing, boundary, cost and outcome, sealed as one entry.

The questions the review is aimed at

Auditor question	What the ledger must produce
Who authorised this action?	The named role and the standing grant in force at execution time
Why this model?	The routing decision, its inputs and the constraints that bound it
Did data cross a boundary?	The classification, the policy version and the crossing or refusal
What did it cost?	Realised cost attributed to the action and the envelope it drew on
Has this record changed?	Tamper-evidence sufficient to answer without trusting the operator

Table 1 — The five questions the ledger exists to answer.

Per action

record granularity

Precondition

the ledger write, relative to completion

Independent

verification path for the reviewer

Why this is not a certification announcement

External review is not the same as certification, and conflating the two would be exactly the behaviour this company was founded to argue against. A certificate states that a control existed at a point in time. This review is aimed at whether the construction survives adversarial reading — whether someone trying to make a record say something convenient can do so undetected. Findings will be published, including the ones that require us to change the design.

“The test of an audit ledger is not whether it satisfies us. It is whether it satisfies someone who assumes we are wrong.”
Dipp AI Research

Sources

Sources for every figure in this article.

Where a number comes from Dipp AI's own analysis or an observed deployment, it is labelled as such and is not presented as an independently audited third-party finding.

Dipp AI Technologies, The Enterprise Superintelligence Report, Vol. I (August 2026)
Dipp AI Technologies, Immutable Audit Ledger review scope (August 2026)

Case studies

Organisations that ran this argument in production.

Modelled reference scenarios with the measurement window, the components enforced and the numbers attached. Each one downloads as a PDF.

Healthcare

A 14-hospital system turned agent activity into evidence its regulator accepts

Prior-authorisation and utilisation-review agents were producing decisions nobody could attribute. Ninety days later every action carried a hashed record naming the clinician whose authority was exercised.

6 days: to assemble a payer audit package
100%: of committed agent actions attributable to a named reviewer

Read the study PDF

Legal Services

A law firm made the instruction, not the output, the thing on the record

When a client asked what the firm had instructed the system to do, the answer was a hashed directive with a partner's name on it.

3: client technology audits passed with no findings
100%: of agent review work traceable to a supervising partner

Read the study PDF

Written by Dipp AI Research.

The desk that edits, sources and dates every piece Dipp AI publishes, and holds the line on what may be claimed. Every figure in this piece carries a source, and corrections are published on the record rather than made quietly.