10220 (2026). Camilla Belich to the Minister of Conservation

10220 (2026). Camilla Belich to the Minister of Conservation: What new artificial intelligence or machine learning systems, tools, or platforms, if any, have been commissioned, contracted, or deployed by any department or entity for which the Minister is responsible since 27 November 2023, listed by system name, supplier, purpose, contract value, and date of commissioning?

Hon Tama Potaka: I am advised that the following projects have been undertaken at the Department of Conservation: Machine vision and image recognition – for detecting animals and filtering camera footage. DOC commonly uses the open-source AI model Megadetector, which is locally hosted on specific DOC computers, and Spyfish Aotearoa, which is a collaborative project with Wildlife.ai. Audio classification – for identifying native species calls. DOC commonly uses the open-source AviaNZ, and open-source Kaytoo, which is locally hosted on DOC computers. Retrieval-Augmented Generation (RAG) – for interpreting complex statutory documents. This is a bespoke solution, hosted on DOC’s Amazon Web Services, using Neo4j and OpenAI. Generative AI – for searching, drafting, summarising, and ideation tasks. Microsoft Copilot and Microsoft Teams Premium are commonly used, along with AI within DOC’s enterprise workflow solution, ServiceNow. Details of specific projects, timeframes and costs are attached.