ADM Review & Escalation Agent for Melbourne In-House Counsel: Catch Undisclosed AI Use in Automated Decisions Before It Reaches Regulators
You run legal for a Melbourne financial services group. The product team has been using an internal model to triage customer eligibility decisions for six months. A complaint lands on your desk: a customer wants to know how the decision was made, whether AI was involved, and whether anyone reviewed it. You don’t have a clean answer. The Australian Solicitors’ Conduct Rules sit behind you — Rule 4 (honesty), Rule 9 (confidentiality), Rule 19 (candour) — and your board wants to know, by Friday, which automated decisions in the last quarter were AI-assisted, which ones were disclosed as such, and which ones a human actually reviewed. The ADM Review & Escalation Agent is built for exactly this question.
The problem
Automated decision-making (ADM) inside Australian organisations has expanded faster than the disclosure and review controls around it. Models triage claims, score applications, flag transactions, and draft customer correspondence — often without a documented record of when AI was involved, what a human reviewer actually changed, or what was disclosed to the affected party. For in-house counsel, this creates a compounding exposure: under the Australian Solicitors’ Conduct Rules, you owe duties of honesty and candour that extend to representations made on behalf of your organisation, including representations about how decisions are made. When a regulator, complainant, or court asks “was AI used, and was that disclosed?”, the answer needs to be reconstructable from records — not from memory.
The threat isn’t a single bad decision. It’s the inability to evidence, at scale, which decisions were AI-assisted, which were reviewed by a qualified human, and which were disclosed to the affected party in terms consistent with your ASCR obligations and your organisation’s stated policies.
What the ADM Review & Escalation Agent does
The ADM Review & Escalation Agent is an Exegesis service shape (catalog reference: 03_Agentic_Solutions/ADM_Review_and_Escalation_Agent.md) that wraps a review and escalation workflow around your existing automated decision pipelines. It is designed for in-house legal teams who need a defensible audit trail covering three questions for every flagged decision: was AI used, was a human reviewer engaged, and was disclosure made.
The agent does not make legal decisions for you. It produces a structured queue of ADM events that require human legal review, applies your configured escalation rules, and writes an immutable record of what was decided, by whom, and on what basis. The output is a per-decision review log and a periodic ADM disclosure register suitable for board reporting and regulator response.
How it works
- Ingest — the agent receives ADM events from your decisioning systems (eligibility, triage, scoring, customer correspondence) along with the model metadata and inputs available at decision time.
- Classify — each event is classified against your ASCR-aligned policy matrix: was AI used, what was the nature of the decision, who is the affected party, and what disclosure obligations attach.
- Route — events meeting escalation thresholds are routed to the nominated in-house reviewer with the supporting context bundled into a single review pane. Lower-risk events are sampled for spot review.
- Record — the reviewer’s determination, the rationale, any disclosure made, and the version of the policy applied are written to an append-only review log.
- Report — the agent produces a periodic ADM register summarising volumes, escalation outcomes, disclosure rates, and outliers — formatted for board, audit committee, and regulator-facing use.
Why this matters in Melbourne
Melbourne in-house counsel work under the Legal Profession Uniform Law and the Legal Profession Uniform Law Australian Solicitors’ Conduct Rules 2015, which have applied in Victoria since 1 July 2015. The ASCR are a statement of professional and ethical obligations that flow from a solicitor’s duty as an officer of the court, and they bind in-house solicitors in the same way they bind private practice — including the duties of honesty (Rule 4) and candour, and the duty to only give effect to client instructions that are lawful and properly authorised. When your “client” is the organisation, representations made by the organisation about how decisions are made — including AI-assisted decisions — become your concern.
The Law Council’s 2026 ASCR review process makes the direction of travel clear: the rules are being actively considered against new disclosure and reporting obligations affecting solicitors, and in-house teams are expected to have the operational controls in place to evidence compliance, not just assert it.
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 —
03_Agentic_Solutions/ADM_Review_and_Escalation_Agent.md - RuleCheck by Exegesis
Join the waitlist
The ADM Review & Escalation Agent is a Tier-3 service shape currently scoping with design partners. Pricing structure (per-decision, per-reviewer monthly, or enterprise licence) will be set based on demand from in-house teams. Join the waitlist and what we hear from you will shape how access opens.