AI Transparency Statement

Purpose

The purpose of this AI Transparency Statement is to provide clear and accessible information about our use of AI systems. We aim to inform the public about how we utilise AI to improve service delivery, enhance operational efficiency, and support our research initiatives. This statement outlines our commitment to transparency, governance, and compliance in the use of AI technologies.

Overview of Use

We use AI systems within our Microsoft 365 environment, including Microsoft 365 Copilot, to support internal operations and business activities in an appropriate and controlled manner.

In line with Australian Government classification approaches, our use of AI can be broadly described as follows:

Usage patterns

• Workplace productivity (e.g. drafting, summarisation, administrative support)
• Analytics and insights (e.g. reviewing information or assisting to identify patterns)

Domains

• Corporate and enabling functions (e.g. administration, finance, HR, internal communications, and operational support)
• Program and research support (e.g. research, analysis, drafting materials, and supporting organisational activities)

These AI capabilities are used to assist personnel by providing timely information, insights, and draft content to support their work. AI is not used to make or implement decisions; all decisions remain the responsibility of appropriately authorised personnel.

Productivity tools: AI tools are used to improve productivity, including through the use of features that assist with the preparation of meeting minutes, summarise content for review, highlight action items, flag emails and similar functionality.

Drafting materials: AI assists in drafting various types of documents and other materials for review, including emails, reports, internal and external communications, to improve efficiency and accuracy.

Search and information retrieval: AI-enabled search capabilities help staff locate, retrieve, and navigate internal information, including policies, guidance, research materials, and organisational knowledge.

Monitoring and Governance

We apply a risk-based approach to the monitoring and governance of AI. Our AI systems are subject to ongoing evaluation to ensure performance, accuracy, and appropriate management of risks, including bias and unintended impacts. Key measures include:

Performance monitoring: Continuous assessment of AI system accuracy and effectiveness.

Bias checks: Regular audits to identify and mitigate any biases in AI algorithms.

Human oversight: AI-generated content is treated as a draft and reviewed by personnel prior to use. AI outputs are used to support analysis and inform decision-making; however, AI does not make or implement decisions.
AgriFutures maintains an internal register of AI systems to support governance, transparency, and assurance activities.

AI systems are used to provide information, analysis, or recommendations to support human judgement. They are not used to make or implement decisions without human involvement.

Compliance

We use AI in compliance with relevant laws and regulations, including the Privacy Act 1988 (Cth), the Australian Privacy Principles (APPs) and other applicable data protection and information security requirements. We are committed to maintaining a high standard of data privacy and security, including by regularly reviewing legal requirements and ethical guidelines to ensure we apply best practice. For more information regarding data privacy, please refer to our Privacy Policy.

View the AgriFutures Australia Privacy Policy

Risk Mitigation

To minimise potential risks associated with AI, we apply a range of controls and safeguards supported by policies and guidelines designed around:

• Fairness: Ensuring that AI systems are designed and deployed in a manner that promotes fairness and equity.
• Security: Implementing robust security measures to protect AI systems from cyber threats and unauthorised access.
• Transparency: Providing clear and accessible information about how AI systems operate.

These policies and guidelines inform staff of appropriate AI usage and to document risk mitigation strategies. They include:

• AgriFutures Artificial Intelligence and AI Chatbot Usage Policy
• Generative AI Guidelines
• ICT Acceptable Use Policy.

AI outputs are subject to human review and validation prior to use in business processes.

AI Use Cases

Microsoft 365 Copilot

AgriFutures Australia is using Microsoft 365 Copilot to improve collaboration across teams and boost individual productivity. Copilot’s AI features help automate routine tasks, search for information across our internal systems, summarise information, assist with data analysis, and generate reports, freeing up staff to focus less on manual work and more on high value tasks, as well as providing timely insights and information to support staff in making more informed decisions.

AgriFutures does not currently deploy AI systems that autonomously make or implement decisions, or that directly affect individuals, without human review and authorisation.

AI Accountable Official

An AI Accountable Official is designated with responsibility for oversight of AI governance, compliance, and implementation of relevant policy requirements.

Contact

If you wish to make any inquiries or requests for further information regarding our AI use and management practices, you can do so by contacting the AI Accountable Official:

Via post
AgriFutures Australia c/o AI Accountable Official
Building 007, Tooma Way, Charles Sturt University
Locked Bag 588
Wagga Wagga NSW 2650
AUSTRALIA

Via phone
+61 2 6923 6900

Via email

Conclusion

We are dedicated to maintaining transparency and accountability in our use of AI, ensuring that AI supports human judgement and does not replace decision-making responsibility.

This statement will be reviewed at least annually, and updated where there are material changes to AgriFutures’ approach to AI.

Review and Update Cycle
This AI Transparency Statement will be reviewed annually, with ad-hoc updates made as necessary to reflect significant changes in our AI systems or approach. The next scheduled review date is December 2026.

Publication and Accessibility
This statement will be published on our public website and can be easily accessed through the global menu/footer. We encourage stakeholders to review this document and provide feedback to help us improve our AI practices.

Ongoing Maintenance
We are committed to the ongoing maintenance of this AI Transparency Statement, ensuring it remains current and reflective of our AI practices.