While many media organisations have had to walk away from funding news operations in regional areas, New Zealand’s Stuff Group has remained committed. To ensure local coverage is as comprehensive as possible with sometimes stretched resources, new solutions were needed to deliver critical local coverage for often remote and underserved communities.  

New Zealand may be a small country, with a population of just over 5 million, but it still has hundreds of municipal government bodies — many in remote areas. With more than 30 national, regional, and community newspaper mastheads across New Zealand, Stuff Group’s Masthead Publishing division is committed to keeping its audiences informed about crucial decisions made by elected officials in their own backyard, no matter how far-flung.

Monitoring local government manually takes significant time and resources, both in increasingly short supply. Generative AI offered a solution.

We developed a local democracy custom-GPT, giving us the option to deep dive into local government affairs and produce fresh, relevant stories for our widespread readership. As one of our news leaders said recently: “I’m old enough to remember when e-mail came to our newsrooms; this feels just as transformative.”

What Democracy.AI does

Democracy.AI gives our reporters the ability to scan, prioritise, and report on hyper-local decision-making documents such as meeting minutes and committee submissions in volume, giving audiences a better picture of what’s happening in their towns.

The tool reviews source material, selects relevant story ideas, and, if required, writes the stories. It gathers information from official papers attached to each request, analyses for relevance and accuracy, and crafts well-written, engaging articles. 

Headlines are optimised for search engines, it quotes accurately from the source data provided, and before writing a story, it will ask how many words the story needs to be. We also gave ourselves extra ears and eyes using AI listening tools to monitor city hall meetings and identify key themes and issues.

Piloting Democracy.AI

We ran a two-month pilot of the tool in one of our regional newsrooms to test the viability of the project. Our Waikato Times newspaper and digital masthead has 11 local authorities in its coverage area, so it was the perfect testing ground. 

We wanted it to be simple, sustainable, effective and well-received by readers. It was.

Democracy. AI gave the power to participate in the decision-making process of local democracy back to the people.

 Waikato Times Editor Jonathan MacKenzie was initially sceptical, but has been amazed by the speed at which his reporters can work when assisted by AI. 

“It’s like having an army of worker bees happily going about their toil, allowing our reporters to focus on more complex and meaningful stories that provide context for readers in their everyday lives,” he said.

“It’s far better for my staff to be out talking to people and digging for stories than stuck behind a desk reading a council agenda. In my experience, the best journalism is done when there’s more time for prospecting in the field, and that’s what AI support affords us.”

The results: Democracy.AI in action

The Waikato Times newsroom used the tool daily, generating significantly more local democracy stories and building a new section on its subscription Web site called Ratepayers’ Roundup, significantly boosting traffic, engagement and reader satisfaction.

Across the pilot period, the Waikato Times increased its story count and digital subscriptions more than doubled, with local government and Ratepayers’ Roundup content being amongst the highest conversion drivers. The subscriber net promoter score for the masthead also grew 19% over the pilot period.

Since the successful pilot, we have rolled the tool out to other regional newsrooms across New Zealand. It’s AI for the people, improving the hyper-local news we publish and giving our journalists more time to explore, write and produce the unique content AI cannot deliver.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .