Book a discovery call

Home / Self-assessment

AUR-WEB-012 · v1.0 · Agentic workforce governance

Is your agentic workforce governed, or just deployed?

Eleven questions, ten minutes. Your organisation already governs its human workforce with job descriptions, managers, reviews and exit processes. This self-assessment measures whether your AI agents get the same discipline.

Answer for your organisation as it is today, not as it is planned to be. If you don't know, that is an answer too: choose the first option. Three questions are marked ADM because they relate to the automated decision-making transparency obligations in Australia's Privacy Act reforms, which commence in December 2026.

Your jurisdiction

0 of 11 answered

01Agent registerADM

Do you have a complete, current register of every AI agent operating in your organisation, including agents staff have built themselves?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

02Mandate

Does every agent have a written mandate: what it is for, what it may do, what it must never do, and what data it can access?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

03Accountable owner

Is a named person (not a team or a department) accountable for each agent?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

04Charter (approval)

Before an agent goes live, are its mandate, owner and review obligations formally approved by someone with authority to do so?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

05Delegation map

Can you say which agents can call on other agents, tools or systems, and what approval each of those connections required?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

06Human checkpointsADM

Are the decisions that always require human sign-off defined for each agent, and is that sign-off actually evidenced?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

07Decision provenanceADM

If an agent made or influenced a decision affecting a person, could you produce a record of what it decided and why?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

08Reviews and drift

Are agents reviewed on a schedule, and automatically re-reviewed when something changes, such as a new capability, a new data source, an incident, or the owner leaving?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

09Third-party assurance

When a vendor or partner supplies an agent that operates on your data, do you assess it before it starts work?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

10Workforce attestation

Could a senior leader sign a statement today that every active agent has an owner, a current mandate and a clean review, with the evidence to back it?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

11Decommissioning

When an agent is retired, or loses its owner, is there a defined offboarding process with credentials revoked and records retained?

No, or I don't knowAd hoc: true for some agents, nothing consistentMostly: consistent practice, gaps in the documentationYes: documented, current, and evidenced

AI governance and conformance consulting for Australian organisations. Anthropic Partner and Microsoft Cloud Partner.

PROUDLY SOUTH AUSTRALIAN

Services

Aureus Accelerate

Company

How we work

Legal

Terms of service

AUR-WEB-012 · v1.0 · Reviewed July 2026

© 2026 Aureus Solutions · ABN 59 608 064 589