Migroo Responsible AI Policy

Overview

‍

Effective date: 13 December 2025 (Version 1.0)
Company: Migroo Pty Ltd (ABN 17 686 506 918)
Business address: Level 3 / 21 William Street, Balaclava VIC 3183, Australia

Status: Public Policy

‍

1. Purpose of this Policy

Migroo is committed to building AI systems that are safe, transparent, fair, and respectful of the people who rely on them. This policy outlines how we design, use, and govern AI across our platform, and the principles that guide our decisions as we continue to develop AI-powered migration tools.

Our goal is simple: ensure that AI improves access to migration information without compromising privacy, fairness, or human dignity.

‍

2. Scope

This policy applies to:

• All AI models and systems used within Migroo
• All data processed for the purpose of AI inference or scoring
• All employees, contractors, and partners involved in AI development, deployment, or review
• Third-party AI tools integrated into the platform (e.g., OpenAI models)

‍

3. Our responsible AI principles

‍

3.1 Privacy by design

Migroo treats privacy as a foundational obligation.

We commit to:

• Collecting only the minimum data necessary
• Never storing user-generated documents longer than our stated retention periods
• Encrypting documents at rest and in transit
• Ensuring all AI processing complies with Australian privacy law and GDPR principles

Migroo does not train AI models on user-uploaded content.
User data is never sold, shared, or used for marketing purposes.

‍

3.2 Mitigation and Fairness

Migration is a deeply human, culturally diverse process.
We commit to reducing algorithmic bias wherever possible.

Migroo will:

• Use diverse datasets and examples during model development
• Regularly audit outputs for signs of unfair or inconsistent scoring
• Seek feedback from users and practitioners to identify potential areas of bias
• Avoid any AI features that could unintentionally discriminate based on race, gender, nationality, relationship structure, or disability

Where AI outputs are complex or sensitive, Migroo provides clear explanations, not black-box decisions.

‍

3.3 Transparency and Explainability

We believe users have the right to understand how AI impacts their experience.

Migroo provides:

• Clear descriptions of what each AI tool does
• Explanations of scoring frameworks (e.g., the four pillars for partner visas)
• Reasons behind feedback or recommendations
• Transparent limitations and disclaimers

We will never position AI as a substitute for legal advice.

‍

3.4 Human Oversight

AI supports people — it does not replace judgment, expertise, or empathy.

Migroo ensures:

• Humans design, evaluate, and monitor all AI tools
• Sensitive outputs (e.g., red flags, inconsistencies) are handled carefully
• Users maintain full decision-making control over their visa applications

Where appropriate, users can escalate to a human migration agent or lawyer for review.

‍

3.5 Safety and Security

Migroo follows industry best practice for AI safety, including:

• Regular model performance testing
• Monitoring for harmful, misleading, or inaccurate outputs
• Guardrails for user content
• Infrastructure hosted on carbon-neutral, secure cloud platforms in Australia

We prioritise user protection in every decision.

‍

3.6 Continuous Improvement

AI governance is never “done.” Migroo commits to continual review and refinement.

We will:

• Update this policy annually
• Monitor emerging regulations (AI Act, privacy reforms, consumer law)
• Improve accuracy, fairness, and accessibility
• Adjust internal processes based on user feedback and expert advice.

‍

4. Data Retention

Migroo applies strict limits on data retention:

• Uploads stored only for operational needs (usually 48 hours)
• Permanent deletion after the retention window
• Metadata retained only for completeness scoring (non-sensitive)

No user data is stored indefinitely or used for training.

‍

5. Third-Party AI Partners

Migroo uses large language models (LLMs) responsibly by:

• Choosing partners with strong safety and privacy standards
• Ensuring all processing complies with local data regulations
• Verifying that third-party AI providers also do not train on Migroo content

We maintain internal controls to ensure all integrations meet Migroo’s ethical requirements.

‍

6. Accountability

Migroo leadership is responsible for enforcing this policy and ensuring all contributors understand their obligations.

Users can provide feedback or raise concerns about AI behaviour at:
info@migroo.com.au

All concerns are taken seriously and investigated promptly.

‍

7. Declaration

Migroo is committed to upholding this policy across all operations.
Our goal is to ensure every contributor — internal or external — experiences respect, inclusion, and fair compensation for their time and expertise.

‍