STRmix Team Artificial Intelligence (AI) Transparency Statement
24 June 2026
The STRmix team at the New Zealand Institute for Public Health and Forensic Science (“PHF Science”) embraces the potential of AI to enhance the efficiency and effectiveness of its operations through human-centred, transparent, deliberate, and careful use. This statement provides a high-level overview of how we use AI.
Guidance on the use of AI by the STRmix team is regularly reviewed to retain relevance as new insights emerge, and to respond to any early signals of risk or harm.
The team adheres to the following guiding principles for the responsible use of AI:
1. Human oversight - AI assists decision making but does not replace human judgment. Critical decisions remain under meaningful human control.
2. Transparency - Be clear when AI is used in external publications, client facing outputs, or research. Disclose AI involvement and maintain openness about its role where AI significantly contributes to analysis or advice.
3. Data protection - Protect personal, confidential, and culturally sensitive data. Follow PHF Science’s privacy, security, and information governance policies at all times.
4. Fairness and inclusivity - Avoid bias and ensure AI use benefits all stakeholders. Regularly review outputs for unintended discrimination or harm.
5. Accountability - Staff and managers share responsibility for ethical AI use. Roles and responsibilities for oversight are clearly defined.
6. Continuous learning - Stay informed about AI capabilities and limitations. Complete required training and share best practices across teams.
7. Innovation with care - Explore new AI applications responsibly, balancing creativity with compliance and risk management. Pilot before full adoption.
8. Cultural respect - Honour Māori Data Sovereignty principles in all AI related activities, ensuring partnership and shared decision making where appropriate.
9. Sustainability - Consider the environmental and resource impacts of AI technologies and adopt practices that minimise harm.
AI tools in use are assessed for compliance with PHF Science’s privacy, security and information and data management policies.
Any use of AI within the STRmix team must ensure that the human user remains the sole responsible author, including but not limited to thorough human verification of all AI output by checking for:
1. Factual accuracy and correctness,
2. Context and completeness, and
3. Alignment with standards.
All use of AI within the STRmix team must also comply with pre-existing contractual obligations, data sovereignty rules, participant agreement, copyright and intellectual property rights.
AI use by the STRmix team can include the following activities:
· Limited parts of the software development or testing process. With regard to commercial software applications (FaSTR™ DNA, STRmix™, DBLR™ and STRmix™ NGS), coding development (including code review and creation or review of unit tests) using AI is only permitted after completion of a full risk assessment, mitigation and approval process. AI has not been used for coding development in any currently available production version of FaSTR™ DNA, STRmix™, DBLR™ or STRmix™ NGS.
· Creating, reviewing/proposing improvements to, formatting or proof-reading of
o documentation
o email or other communications
o media or presentations
o test scripts and plans
o images or music
· Reviewing trends in data and generating insights
· Translating text
· Summarising meetings, chats and documents
· Administrative tasks.
Where AI has significantly contributed to analysis, authoring or advice for external parties, documents must clearly state that AI has been used to assist in their preparation, with subsequent human verification for factual accuracy, context and completeness, and alignment with standards.
The following are examples where the use of AI is not permitted:
1. Writing or revising material in a way that the human author loses ownership and is no longer able to explain, defend and maintain the content; and
2. The end-to-end use of AI in a creation process, where the human user has no input into the creation or verification of the output.
The incorporation of AI functions into the STRmix team’s software applications is governed by a separate risk assessment and decision-making process. AI functionality is currently limited to the optional use of Artificial Neural Networks (ANNs) in FaSTR™ DNA. There is no AI functionality in STRmix™, DBLR™ or STRmix™ NGS.
Definition
· “Artificial Intelligence”, or “AI” is the simulation of human intelligence processes by machines, especially computer systems, including AI Agents. These processes include learning (the acquisition of information and rules for using the information), reasoning (using rules to reach approximate or definite conclusions), generation (using learned rules to create data), and self-correction. AI systems can analyse information, identify patterns, generate content, or support decisions in ways that simulate aspects of human reasoning. For the avoidance of doubt, AI refers to AI tools or systems that materially influence work outputs or decisions and does not include common productivity features embedded in standard workplace software.
You can contact us about this Transparency Statement by:
Last modified: