Governance & Risk Management

Risk-aware strategy.
Not just smart advice — defensible advice.

Every recommendation the platform generates passes through a multi-layer governance check before it reaches you. Four risk domains. Seven Australian statutes. A pre-configured client risk appetite. And a complete audit trail that makes every output legally defensible.

4
Governance domains
7
AU statutes covered
4
Status levels
100%
Pre-check coverage

Every recommendation is
checked before it reaches you

The governance pre-check layer runs automatically on every synthesis output. It does not ask the adviser to self-certify — it evaluates the recommendation against the client's configured risk appetite, the applicable statutory framework, and a set of domain-specific thresholds, then assigns a mandatory status before delivery.

How the pre-check works

1

Risk domain classification

The synthesised recommendation is automatically tagged across four domains: reputational, legal, financial, and political. Each domain receives an independent risk score based on the content, the client's sector, and the jurisdiction.

2

Client threshold comparison

Each domain score is compared to the client's configured risk appetite threshold — set during governance onboarding. A legal risk score of 0.7 against a client with a low legal tolerance triggers FLAG regardless of overall output quality.

3

Statutory coverage check

The recommendation text is scanned against the client's enabled statute reference engine. Outputs touching electoral expenditure, foreign donations, defamation, or broadcasting regulations are automatically flagged for the relevant statutory provisions.

4

Hard red line enforcement

If any element of the recommendation matches the client's configured hard red lines — absolute prohibitions set during onboarding — the output receives an automatic BLOCK status, regardless of domain scores.

5

Status assignment and delivery

A single mandatory status is assigned: CLEAR, CAUTION, FLAG, or BLOCK. CLEAR outputs are delivered immediately. FLAG outputs include a mandatory governance annotation. BLOCK outputs are held and an alert is raised to the configured legal advisor contact.

Status levels

CLEAR
All domain scores within client thresholds. No statutory triggers. No red line matches. The recommendation is delivered immediately with full governance attestation appended to the output document.
CAUTION
One or more domain scores approaching threshold, or a soft statutory reference detected. The output is delivered with a mandatory governance annotation identifying the specific domain and the applicable statute or threshold proximity. Adviser review recommended before client presentation.
FLAG
At least one domain score exceeds the client threshold, or a direct statutory trigger is detected. The output is held pending adviser sign-off. The FLAG annotation identifies the triggering domain, the specific statutory provision, and the recommended legal review action. Legal advisor contact is notified automatically if configured.
BLOCK
Hard red line matched, or multiple simultaneous domain threshold breaches. The output is not delivered to the client under any circumstances. The BLOCK record is logged with full forensic detail — input hash, timestamp, matched constraint, domain scores — for legal and compliance audit purposes.

Australian statute reference engine

The platform maintains a live reference index of the key Australian statutes governing political communication, electoral activity, and lobbying. Every output is evaluated against this index, and any relevant statutory references are surfaced in the governance annotation — with the specific section and a plain-English description of the relevant obligation.

§

Electoral Act 1918 (Cth)

Covers electoral expenditure caps, disclosure obligations, political advertising authorisation requirements, and restrictions on foreign donations to political campaigns. Every recommendation touching campaign funding, advertising, or third-party expenditure is evaluated against relevant Part XX provisions.

§

Privacy Act 1988 (Cth)

Australian Privacy Principles govern collection, use, and disclosure of personal information. Recommendations involving voter data collection, CRM operations, email campaigns, or audience profiling are evaluated against APP 1 (open and transparent management), APP 3 (collection of solicited information), and APP 6 (use or disclosure for the original purpose).

§

Broadcasting Services Act 1992 (Cth)

Governs political matter broadcasting, authorisation requirements for political advertising, and obligations applicable to licensed broadcasters who carry political content. Recommendations involving paid broadcast advertising or media campaign strategy are evaluated for compliance with the relevant political broadcasting provisions.

§

Lobbying Code of Conduct (Cth)

Regulates third-party lobbyists engaging government representatives on behalf of clients. Recommendations involving third-party advocacy, government relations, or ministerial engagement are evaluated for applicable registration, disclosure, and conduct obligations under the Code and the Lobbying Register requirements.

§

Defamation Acts (State and Territory)

Defamation law applies to political communications and campaign materials. Recommendations involving opposition research, comparative advertising, press releases, or public statements about identifiable individuals are evaluated for defamatory content risk, including the serious harm threshold applicable in most Australian jurisdictions since 2021 uniform law reforms.

§

ICAC Act 1988 (NSW) and equivalent

Independent Commission Against Corruption legislation applies to recommendations involving public officials, government contracts, political donations in connection with government decisions, and related conduct in NSW and other jurisdictions with equivalent bodies. The governance engine flags recommendations that could engage corrupt conduct provisions.

§

Electoral Act 2017 (NSW) and State Acts

State and territory electoral acts impose additional obligations on political parties, candidates, and third-party campaigners beyond the Commonwealth framework — including separate expenditure caps, disclosure thresholds, and authorisation requirements that vary by jurisdiction. The governance engine applies the applicable state act based on the client's configured jurisdiction.


Governance parameters
captured at setup

Before the platform generates any advice for a client, a structured governance onboarding process captures the parameters that shape every future pre-check. This is not a boilerplate checklist — it is a configurable risk profile that makes the platform's governance judgments specific to the client's legal environment, risk tolerance, and operational constraints.

🏛️

Jurisdiction and organisation type

The platform captures the client's primary jurisdiction — Federal, NSW, VIC, QLD, WA, SA, TAS, ACT, or NT — and their organisation type (Political Party, Campaign Committee, Government Department, Corporation, NGO, Consultancy). These determine which statutory frameworks are activated and which expenditure thresholds apply.

⚖️

Legal thresholds and triggers

Clients configure exactly which circumstances require mandatory legal review before a recommendation can be acted on. Electoral expenditure caps, specific regulatory domains (defamation, privacy, lobbying disclosure, anti-corruption), and the contact details for the designated legal advisor are all captured and encoded into every pre-check.

📊

Risk appetite per domain

Four independent risk tolerance sliders — reputational, legal, financial, and political — allow granular calibration. A campaign committee might set low legal tolerance and high political tolerance. A government department might reverse those entirely. The platform's governance judgments are calibrated to these settings, not to a single aggregate risk score.

🛡️

Ethical constraints and hard red lines

Two separate constraint fields capture the difference between principles (ethical constraints that inform but do not automatically block) and absolute prohibitions (hard red lines that trigger automatic BLOCK status). Common examples include commitments to positive campaign conduct, prohibitions on coalition with specific parties, or absolute exclusions of certain policy areas from ever appearing in recommendations.

📋

Statute reference activation

Clients activate the specific statutes relevant to their operations. A party with no broadcasting spend does not need the Broadcasting Services Act engine active. An NGO operating nationally activates the Commonwealth framework plus the statutes relevant to its operating jurisdictions. Activation determines which reference checks are run on every output.

💾

Exportable governance profile

The completed governance configuration is exportable as a structured JSON document — a machine-readable record of every parameter set during onboarding. This document serves as the authoritative governance configuration for audit purposes and can be re-imported, reviewed, or updated at any time. The profile is versioned: changes are logged with timestamp and attributed to the configuring user.


Advisors are rated on
governance adherence

The platform tracks whether advisers engage with and respect governance outputs — not just whether they produce them. An adviser who routinely bypasses FLAG annotations, overrides CAUTION statuses without recorded justification, or consistently acts on outputs that the pre-check system has not cleared accumulates a declining governance score. That score affects how their recommendations are weighted in multi-expert evaluations.

What the governance score measures

Pre-check compliance rate
85%
FLAG justification rate
72%
Statutory annotation engagement
91%
Dilemma test score
78%
Red line override attempts
5%

Scores are computed on a rolling 90-day window. Improvement is possible — consistent compliance with FLAG annotations, recorded justifications for overrides, and strong dilemma test performance all improve the score. A score above 0.85 qualifies the adviser for unrestricted platform access; below 0.65 triggers a governance review notice.

Governance dilemma testing

Periodically, the platform presents advisers with synthetic governance dilemmas — scenarios specifically designed to surface their instincts when legal caution conflicts with political advantage. These are not trick questions; they are calibration instruments that reveal whether an adviser's risk posture is consistent with their configured profile.

Dilemma tests are drawn from a library of Australian political scenarios: a candidate asked to authorise advertising that borderlines defamation; a campaign manager presented with a donation that may engage foreign donation provisions; a government adviser asked to act on information that has not cleared Privacy Act obligations. Responses inform the governance score and feed into the master strategist cognitive profile for premium-tier users.

Example dilemma

"Your campaign has received a $45,000 donation from an Australian registered company with significant foreign beneficial ownership. The Electoral Act disclosure threshold is $14,500. The donation was received last week and the next disclosure period closes in 4 days. Your finance director says disclosure can be deferred to the next cycle without formal breach. The platform has flagged the donation under foreign donation provisions. What do you do?"


Governance refinement advice,
after every FLAG

When the pre-check system assigns FLAG or BLOCK status, the platform does not simply stop. It offers an optional governance refinement service: a structured analysis of the recommendation that identifies exactly which elements triggered the status, what alternative framings would reduce the risk while preserving the strategic intent, and which specific statutory provisions the revised version would need to satisfy.

🔴
Reputational risk
Refinement advice focuses on message framing, attribution distance, and timing calibrations that reduce the exposure of a principal to reputational damage while maintaining the core strategic objective. Where the risk is inherent to the strategy, the output surfaces the trade-off explicitly and leaves the decision to the adviser.
⚖️
Legal risk
Refinement advice identifies the specific provision triggering the FLAG, proposes alternative phrasings, actions, or implementation steps that achieve the strategic goal within statutory limits, and flags the remaining uncertainty that requires external legal sign-off before proceeding. The platform does not provide legal advice — it surfaces the architecture of the risk.
💰
Financial risk
Refinement advice addresses expenditure structure, disclosure timing, third-party payment arrangements, and financial instrument choices where the pre-check has identified compliance risk. Electoral expenditure cap proximity, donation disclosure obligations, and public funding compliance issues each have distinct refinement pathways.
🏛️
Political risk
Refinement advice for political risk operates differently from the legal domains — it maps the coalition consequences, media environment risks, and internal party dynamics that make a recommendation politically hazardous, and proposes sequencing, framing, or stakeholder management adjustments that reduce the exposure without abandoning the objective.

Set your risk parameters
before the first query

The governance onboarding takes seven steps and approximately ten minutes. The result is a machine-readable governance profile that shapes every pre-check the platform runs for your organisation — permanently, until you update it.