When elected representatives lack the ability to scrutinise complex proposals effectively, they become dependent on staff recommendations rather than exercising robust independent judgement. Photo / Warren Buckland
When elected representatives lack the ability to scrutinise complex proposals effectively, they become dependent on staff recommendations rather than exercising robust independent judgement. Photo / Warren Buckland
Opinion by Nick Clark
Nick Clark is a Senior Fellow at The New Zealand Initiative (www.nzinitiative.org.nz).
THE FACTS
Councillors face overwhelming workloads, struggling to scrutinise complex proposals effectively, leading to dependency on staff recommendations.
AI could enhance councillors’ analytical capabilities, providing summaries and insights to improve decision-making and oversight.
Pilot programmes and training are needed to implement AI tools effectively, ensuring councillors remain in control.
Picture this scenario: It is a Friday afternoon. A councillor receives a 600-page agenda for Tuesday’s meeting. Among the technical reports, planning documents and financial statements lies a critical decision about her community’s future.
She has the weekend to make sense of everything while juggling her day joband family commitments. By Monday evening, she faces a stark choice: pull an all-nighter and hope some of it sticks or arrive unprepared for a meeting where fulltime professional staff hold all the analytical cards.
This plays out in council chambers across New Zealand, creating an imbalance that undermines local democracy. When elected representatives lack the ability to scrutinise complex proposals effectively, they become dependent on staff recommendations rather than exercising robust independent judgement.
Neglected critical infrastructure, sluggish services, and hefty rate increases are all too common. Headlines are grabbed by costly “nice-to-haves” like new convention centres, town halls, hotels, festivals, and branding campaigns. Councillors report feeling overwhelmed, under pressure and inundated with complaints. In this environment, attracting good people to serve becomes harder and harder.
My latest research for the New Zealand Initiative reveals how artificial intelligence could help restore balance and improve the lot of elected representatives. Not by replacing human judgement, but by augmenting councillors’ analytical capabilities.
Providing each councillor with dedicated research staff would be prohibitively expensive. AI offers the same analytical capabilities - rapid information processing, risk identification, and community feedback synthesis - at a manageable cost.
Consider how AI assistance transforms that Friday afternoon scenario. Within minutes, the councillor receives plain-language summaries highlighting key decision points and financial implications. The system flags inconsistencies with previous council positions, identifies similar cases from other jurisdictions, and analyses the proposal from multiple expert perspectives.
She arrives at Tuesday’s meeting having identified subtle flaws that might otherwise escape notice, armed with specific questions that enhance rather than undermine scrutiny.
AI in local government is not theoretical. Several New Zealand councils have been experimenting.
Hutt City Council reports staff participants saving an average of 38 minutes daily through AI tools - equivalent to 20 working days annually. Nelson City Council used AI to process Long Term Plan submissions, performing sentiment analysis and thematic categorisation that would have taken weeks manually. Auckland Council has launched “Ask Auckland Council”, a digital assistant with multilingual capabilities.
Yet, these initiatives focus primarily on operational efficiency and service delivery.
What remains largely unexplored is AI’s potential to strengthen the democratic heart of local government: the capacity of elected representatives to provide adequate oversight.
The information asymmetry problem has intensified as councils face increasingly complex regulatory environments. The Resource Management Act comprises over 800 pages and is constantly being amended. Add the Local Government Act, Land Transport Management Act, Building Act, and Water Services Act, and the compliance landscape is a dense labyrinth.
Elected members typically have minimal support beyond their own time and determination. Photo / Bevan Conley
Staff work fulltime in specialist areas, building deep institutional knowledge over years or decades. They control information flow, determine which questions reports address, and frame policy options in ways that significantly influence outcomes.
The resource gap is equally stark. Professional staff have access to research teams, consultant reports and expert networks. Elected members typically have minimal support beyond their own time and determination.
For part-time councillors balancing other commitments, the constraints become overwhelming. This is not a criticism of council staff. It is simply the reality of the system.
New Zealanders remain deeply sceptical of AI technologies - more so than citizens of any other surveyed country. This public wariness creates implementation challenges that councils must carefully navigate through public engagement and ethical guidelines.
AI tools designed explicitly for elected representatives could help address these structural disadvantages. But humans must remain in control. When communities face difficult trade-offs between competing priorities, the decisions involve political and moral reasoning that should not be outsourced to machines.
Critically, AI should not become a crutch used unthinkingly by councillors to substitute for knowledge and experience. Training will be needed to avoid uncritical reliance on AI outputs.
Pilot programmes, shared learning networks, and capability development will also be needed so councils can roll out AI tools effectively and efficiently.
In an era of growing information complexity and rising community expectations, the status quo is becoming untenable – a governance system where those elected to lead cannot increasingly do so effectively.
Applied thoughtfully, AI can help restore balance between elected representatives and professional staff, attract and retain good people, and strengthen rather than weaken democratic accountability.
Ultimately, this could make for local government that works for all of us. And if it saves us from a few white elephants and double-digit rate increases along the way? Even better.
