AI Takes on Royal Duties: New Zealand’s Honours System Gets a Tech Upgrade
The New Zealand government is quietly experimenting with a powerful new tool: generative AI. Specifically, the Department of Prime Minister and Cabinet (DPMC) is trialling the use of AI, through a system called Paerata, to draft citations for the New Year’s Honours list. This move, revealed through an Official Information Act request, signals a broader trend – the integration of AI into traditionally human-driven processes, even those steeped in tradition and requiring nuanced judgment.
Beyond Drafting: The Expanding Role of AI in Government
This isn’t simply about automating a writing task. Paerata, developed internally by the Treasury and its Central Agencies Shared Services (CASS) group, represents a strategic investment in AI capabilities within the public sector. The system operates within a secure cloud environment, addressing key concerns around data privacy – a critical factor when dealing with sensitive personal information contained in nomination documents. The DPMC’s application for an exemption to standard AI usage policies highlights the careful consideration being given to ethical and security implications.
The potential benefits are clear: increased efficiency, reduced workload for staff, and potentially, more consistent and concise citations. However, this is just the tip of the iceberg. We’re likely to see AI increasingly used for tasks like policy analysis, data interpretation, and even public consultation. For example, the UK government has been exploring AI-powered chatbots to handle citizen inquiries, freeing up human agents for more complex issues. A recent report by McKinsey estimates that AI could automate up to 30% of government work globally.
The Human-in-the-Loop Approach: Maintaining Trust and Accuracy
Crucially, the DPMC isn’t handing over the entire process to AI. The current plan involves AI-generated drafts being meticulously reviewed by staff members and then sent to the honouree for proofing. This “human-in-the-loop” approach is vital. Generative AI, while powerful, is prone to errors, biases, and a lack of contextual understanding. Relying solely on AI-generated content could lead to inaccuracies, misrepresentations, or even offensive language.
This mirrors best practices being adopted across various industries. In healthcare, AI is used to assist with diagnosis, but doctors always retain final decision-making authority. In finance, AI algorithms detect fraudulent transactions, but human analysts investigate and confirm suspicious activity. The key is to leverage AI’s strengths – speed and efficiency – while preserving human oversight and judgment.
Privacy Concerns and the Future of Data Handling
The DPMC’s privacy impact assessment rates the risk of harm as “negligible,” but the use of AI to process personal information will inevitably raise concerns. The assurance that Paerata doesn’t learn from the data and doesn’t share it outside the department is reassuring, but ongoing vigilance is essential. As AI models become more sophisticated, the potential for data leakage or misuse increases.
This case highlights the need for robust data governance frameworks and clear ethical guidelines for AI deployment in the public sector. Governments worldwide are grappling with these issues, with the EU’s AI Act setting a new global standard for regulating AI technologies. New Zealand will need to adapt its own policies to ensure responsible AI innovation.
Looking Ahead: AI and the Evolution of Public Service
The DPMC’s experiment with Paerata is a microcosm of a larger transformation underway in the public sector. AI is poised to reshape how governments operate, deliver services, and engage with citizens. We can expect to see:
- Personalized Services: AI-powered chatbots and virtual assistants providing tailored support to citizens.
- Predictive Policing: Algorithms analyzing crime data to identify hotspots and allocate resources more effectively (though this raises ethical concerns about bias).
- Automated Regulatory Compliance: AI systems monitoring businesses for compliance with regulations, reducing the burden on regulators.
- Enhanced Policy Making: AI tools analyzing vast datasets to inform policy decisions and predict the impact of different interventions.
However, successful implementation will require careful planning, investment in skills development, and a commitment to transparency and accountability. The New Zealand government’s cautious approach with Paerata – prioritizing security, privacy, and human oversight – offers a valuable lesson for others navigating this complex landscape.
Frequently Asked Questions (FAQ)
- What is Paerata?
- Paerata is a generative AI tool developed for internal government use in New Zealand, designed to synthesize information and create concise summaries.
- Will AI replace staff in the Honours Unit?
- No, the current plan is for AI to assist with drafting citations, but all drafts will be reviewed by staff and proofed by the honouree.
- What data is being used by Paerata?
- Paerata uses information from nomination forms, supporting letters, and publicly available data related to the honouree’s achievements.
- Is the personal information secure?
- Yes, Paerata operates within a secure cloud environment and is designed not to learn from or retain the personal information it processes.
Want to learn more about the future of AI in government? Explore our articles on digital transformation in the public sector and the ethical implications of artificial intelligence. Share your thoughts in the comments below!
Related reading