AI Transparency Notice

1. Purpose of this notice

The Australian Government's AI Ethics Principles and OAIC guidance on privacy and AI recommend that organisations clearly identify when AI is being used, explain how it works, and be transparent about its limitations. This notice fulfils that commitment for oiva.com.au and describes our design principles for the production Oiva platform.

2. The Action Pack demonstration

The Action Pack preview on oiva.com.au is a demonstration only. It is not a live AI system. It displays pre-authored, editorially reviewed responses to illustrative compliance scenarios. All demonstration responses are manually written by Oiva staff, fact-checked against official sources, and accompanied by the source they are drawn from.

The demonstration does not process or respond to free-text input, connect to any AI language model, access data about your service, store any information from your interaction, or make compliance determinations or provide legal advice. Do not enter personal information about children, families or staff into any demonstration on this site.

3. AI controls summary

The following controls apply to the production Oiva platform currently in development:

4. AI in the production platform

Human review is mandatory. Every compliance-affecting action in Oiva requires a person to review and approve before it is submitted, escalated or finalised. AI drafts. Your team decides. There is no code path that bypasses this.

PII minimisation. A PII detection and minimisation gate runs before any data reaches an AI model. Children's names, family contact details and identifying information are detected and removed or pseudonymised before AI processing. This is an architectural control, not a policy.

Source grounding. Oiva's AI is grounded in ECEC-specific regulatory content. The platform does not generate compliance guidance without a citable official source.

No training on customer data. Customer data is not used to train foundation AI models. Final contractual confirmation with model providers will be published in our Trust Centre before the platform launches.

No biometric processing. Oiva does not process biometric information at any point in its current or planned workflows.

5. What AI model is used

The production Oiva platform is designed to use Amazon Bedrock (ap-southeast-2) with foundation models operating within Australian AWS infrastructure. All AI processing is configured to occur within AWS ap-southeast-2 or ap-southeast-4. This website's demonstration does not use any AI model.

6. Automated decisions and your rights

Oiva does not make fully automated decisions that affect individuals' rights, entitlements or compliance status. All compliance-affecting outputs require human approval. From 10 December 2026, new obligations under APP 1.7-1.9 will require disclosure of substantially automated decision processes. This notice will be updated before that date.

7. Australian AI Ethics Framework alignment

Oiva is designed to support alignment with the Australian AI Ethics Principles and the Voluntary AI Safety Standard. Oiva does not claim formal certification against these standards at this time. We use the language “designed to support alignment with” rather than “certified” or “compliant”.

8. Questions and concerns

Contact privacy@oiva.com.au. We will respond within 10 business days.