AI To Streamline Building Consent Process

Author: Ben O'Connell
AI To Streamline Building Consent Process

There were 33,530 new homes consented in Aotearoa in the year ended May 2025, down 3.8% compared with the year ended May 2024, according to new figures released by Stats NZ.

“The record for the annual number of new homes consented was 51,015 in the year ended May 2022. While consent numbers fell sharply after that peak, they have levelled out over the past year,” economic indicators spokesperson Michelle Feyen said.

In the year ended May 2025, 17,852 multi-unit homes were consented, an 8.6% decrease compared to the year ended May 2024. There were 15,678 stand-alone houses consented, representing a 2.4% increase over the same period.

The seasonally adjusted number of new homes consented in May 2025 rose 10% compared with April 2025. This follows a seasonally adjusted drop of 15% in April 2025.

The seasonally adjusted number of new stand-alone houses consented rose 3.5% in May 2025. This follows a 5.4% fall in April 2025.

How AI Is Impacting Consent Process

With home consent figures stabilising after earlier declines, the focus is shifting to how digital tools can streamline the consent process and support sector efficiency.

A new BRANZ research project explores this, focusing on how AI can automate processes that typically require extensive manual documentation.

The research asks building consent officers about the pain points they experience in processing applications, and where they think AI could eliminate them. Missing documents and inaccurate information are two pain points that have already been identified.

Currently, pre-submission checks by councils only confirm if required documents are present, not whether they’re complete or accurate. BCOs believe that AI could improve this vetting by identifying missing elements, such as site plans, boundaries, and construction details.

BCOs are also interested in AI tools that can quickly interpret specifications, standards, and compliance data, such as NZS 3604 lintel sizes or material compatibility under E2/AS1.

Another key area is communication: BCOs want AI to help draft better Requests for Information (RFIs) by simplifying technical language, improving grammar, and referencing relevant Building Code clauses to make requirements clearer for applicants.

“That’s a challenge in itself,” one BCO said on crafting high-quality RFI letters. “How do you communicate a reasonably technical thing in a way that can get you the answer that you need?”

BRANZ researchers are currently gathering more use cases from BCOs. They will next consult with applicants to determine which AI applications are feasible to trial and implement in the building consent process.