ADM Review & Escalation Agent for Perth In-House Counsel: Catch Undisclosed AI Use Before It Becomes a Candour Problem

You’re the in-house counsel for a Perth business that runs automated or semi-automated decisions — credit assessments, claims triage, vendor screening, internal HR actions. A complaint lands on your desk. The reviewer flagged that “the system” decided. You don’t yet know which steps were model-assisted, which steps were rule-based, whether anyone disclosed the AI component to the affected person, or whether your external solicitors were briefed on the full decision trail before they responded. Under the Legal Profession Uniform Law Australian Solicitors’ Conduct Rules 2015 (in force in Western Australia since 1 July 2022), the candour and honesty obligations attach to anything you or external counsel put on the record. The ADM Review & Escalation Agent is built to make the AI-use facts visible before they become a disclosure problem.

The problem

Administrative decision-making (ADM) inside Australian organisations now routinely blends rule engines, model outputs, and human sign-off. The people closest to the system know which components touched a given decision; the people who have to defend that decision — you, and the external solicitors you brief — often don’t, until a complaint, a regulator letter, or a discovery request forces a reconstruction.

The Australian Solicitors’ Conduct Rules are a statement of professional and ethical obligations binding on solicitors in WA via the Legal Profession Uniform Law. The duty of candour to the court (Rule 19), the duty of honesty in dealings with third parties (Rule 22), and the duty not to make misleading statements (Rule 34) all sit underneath any statement made about how a decision was reached. If a model-generated recommendation drove a decision and that fact is omitted from a position statement, an affidavit, or a response to a regulator, the omission is the solicitor’s problem — not the system vendor’s.

For in-house counsel, the practical risk is more specific. You sign off on positions before they go out. You instruct external solicitors. If your instructions don’t accurately describe AI involvement in the underlying decision, the disclosure problem propagates downstream into documents you don’t draft yourself.

What the ADM Review & Escalation Agent does

The ADM Review & Escalation Agent is a structured workflow for in-house counsel to review an administrative decision, identify which components were AI-assisted, classify the decision under an internal escalation matrix, and produce a brief that external counsel and regulators can rely on as an accurate account of AI use.

It produces, per decision under review:

• A component map: rule-based steps, model-assisted steps, human sign-off points, and the data each step consumed
• An AI-use disclosure status: what was disclosed to the affected person at the time, what was disclosed in any subsequent correspondence, and what is currently in the file
• A candour-risk classification flagging statements already made (internally or externally) that may be incomplete or inconsistent with the component map
• An escalation recommendation: keep in-house, brief external counsel with the component map attached, or escalate to executive/board level
• A reviewable record suitable for sitting alongside the matter file and for use when briefing external solicitors

How it works

1. Intake. You submit the decision artefact (decision letter, internal memo, or system export) and identify the systems involved.
2. Component decomposition. The agent walks the decision pathway and tags each step as rule-based, model-assisted, or human, drawing on system documentation you supply.
3. Disclosure reconciliation. The agent compares what the affected person and any external party have been told against the component map and flags gaps.
4. Candour-risk classification. Each existing statement is rated for consistency with the reconstructed decision pathway.
5. Escalation packet. The output is a structured brief — component map, disclosure status, risk classification, and a recommended next action — ready to send to external counsel or to retain on file.

Why this matters in Perth

Western Australia came into the Legal Profession Uniform Law regime on 1 July 2022, and the Legal Profession Uniform Law Australian Solicitors’ Conduct Rules 2015 apply to WA solicitors from that date. In-house counsel admitted in WA are bound by the ASCR in the same way as private practitioners. Perth in-house teams sit across resources, financial services, utilities, and state-government-adjacent organisations — sectors where administrative decisions are made at volume, are increasingly model-assisted, and attract scrutiny from federal regulators, the State Administrative Tribunal, and the Information Commissioner. The cost of getting AI-use disclosure wrong is not theoretical: it is a candour problem that lands on the solicitor who signed the document, regardless of where in the organisation the AI component was introduced.

Sources

• Law Council of Australia — Australian Solicitors’ Conduct Rules: https://lawcouncil.au/policy-agenda/regulation-of-the-profession-and-ethics/australian-solicitors-conduct-rules

Exegesis capability references:

• ADM Review & Escalation Agent spec
• RuleCheck by Exegesis (open-source verifier)

Join the waitlist

Join the waitlist — be the first to know when access opens for Perth in-house legal teams

The ADM Review & Escalation Agent is in scoping. We’re working with in-house counsel in Perth to shape the component-map taxonomy, the escalation matrix, and the brief format so that it slots into how your team already works with external counsel. Join the waitlist and we’ll come to you with scoping questions before we set pricing.

Schema