Anglican AI policy work includes Indigenous knowledge and workers

A work programme, not a finished policy

AI policy work in the Anglican Church in Aotearoa, New Zealand and Polynesia has begun with a remit that extends beyond what a minister may ask a chatbot to write. An October 20 Anglican Taonga report says the team behind the church’s AI motion met online the previous week to start its work. The report looks toward the church’s 2028 General Synod. Source: anglicantaonga.org.nz

The decision itself is older. Signed minutes from August 20 show General Synod supporting Tim Cross and Bishop Anne van Gend, or an agreed nominee, in preparing terms of reference for an AI working group. The proposed group would report to the General Synod Standing Committee and provide a report to General Synod in 2027/28. That is a mandate to develop the work, not evidence that a final AI policy is already operating. Source: anglican.org.nz

Whose knowledge and whose work?

The adopted motion includes children and future generations; intellectual property and its sovereignty, including “ancient intelligence”; the environment affected by AI systems and infrastructure; and the welfare of people working in the AI sector. It also seeks theological reflection and engagement with work elsewhere in the Anglican Communion. Source: anglican.org.nz

Taonga connects the ancient-intelligence language to Indigenous Pacific knowledge systems, describing a contribution from Polynesia’s Anglicare coordinator, Namuaimana (Mana) Matalavea. This gives the phrase a specific community context; it is not another name for a machine’s capabilities. Source: anglicantaonga.org.nz

The church’s structure matters here. Its constitution provides for three partners to order their affairs within their own cultural contexts: Tikanga Māori, Tikanga Pākehā and Tikanga Pasefika. This is therefore not simply a national committee speaking for one undifferentiated community. Source: anglican.org.nz

What church leaders can learn now

For Christian leaders considering parish AI rules, the useful distinction is between permission to use a tool and responsibility for the people and knowledge involved. A policy confined to checking a generated sermon could leave separate questions unanswered: who has authority to share community knowledge, what information should stay out of a system, and whose work supports the service?

Those are our practical questions, not rules this church has already adopted. The published mandate identifies subjects for deliberation; it does not establish that particular safeguards work or that communities have approved particular uses.

The sources reviewed here do not include final terms of reference or recommendations. Those documents, and the people assigned responsibility, will be the next useful evidence for assessing the work. Churches can learn from the breadth of this agenda without treating the progress report as a completed model to copy.