AI and privacy: what NZ businesses need to know about using AI with personal data
Using AI to process personal information from third-party sources creates unique privacy challenges. Here is the current legal position.
Artificial intelligence is increasingly being used by NZ businesses to process personal information — from automated credit scoring to AI-driven fraud detection. When that personal information comes from third-party sources, the intersection of AI and privacy law creates new compliance challenges.
IPP3A and AI
If your AI system processes personal information collected from a third-party source, IPP3A still applies. The fact that the processing is automated does not remove the notification obligation. In fact, it may strengthen it — the OPC has indicated that automated processing of personal information warrants more transparency, not less.
Transparency obligations
When an AI system makes decisions that affect individuals based on third-party data, best practice (and likely future law) requires:
- Informing the individual that their data was collected from a third party.
- Informing them that the data was processed by an AI system.
- Explaining the general logic of the AI system.
- Providing a mechanism to challenge the decision.
Current law vs future law
The Privacy Act 2020 does not specifically regulate AI. However, the existing principles — particularly IPP3A (notification), IPP6 (access), and IPP8 (accuracy) — apply to AI systems that process personal information. The 2026 review is expected to introduce AI-specific provisions.
DEIS and AI
DEIS allows you to tag lookups that are processed by AI systems, creating a specific compliance record that demonstrates transparency. As the law evolves, DEIS will add AI-specific compliance pathways.
DEIS now integrates with illion: three credit bureaus, one compliance platform
NextLogistics and freight: privacy compliance for identity verification and proof of delivery