Journalism in New Zealand is under structural pressure — smaller newsrooms, faster publishing cycles, more platforms to feed, and ongoing pressure to do more with less. AI can meaningfully help with the volume and speed demands of modern journalism, but the craft of reporting — building sources, asking the right questions, and exercising editorial judgment — remains irreducibly human.
Where AI Adds Real Value for Journalists
1. Research and Background Preparation
Before an interview or investigation, AI can help you rapidly synthesise background — summarising a company’s history, a politician’s voting record, the context of a policy debate, or the background on a technical topic you’re covering for the first time. What previously took an hour of reading can take fifteen minutes.
Verification is essential: AI can hallucinate facts, misattribute quotes, and present outdated information confidently. Any fact generated by AI must be verified against primary sources before publication. Use AI for orientation and research efficiency — not as a source.
2. Transcription and Interview Processing
AI transcription tools can convert recorded interviews to text faster and more accurately than manual transcription. Tools like Otter.ai, Fireflies, and similar services can produce workable transcripts in minutes. Accuracy still varies, particularly with NZ accents, technical terminology, and multiple speakers — always review transcripts against your recording before quoting.
Source protection critical point: If you’re transcribing interviews with sensitive sources, consider very carefully which tools process that audio and where the data goes. Consumer transcription services send audio to cloud servers. For sensitive sources, local transcription tools or manual transcription may be necessary to protect source identities.
3. Drafting and Structure
AI can help with the mechanical aspects of drafting — producing a first structure from your notes, writing a headline and standfirst, or drafting a summary paragraph. For routine news stories (council decisions, earnings releases, sports results), AI-assisted drafting can significantly speed production.
For features, investigations, and stories requiring voice and judgment, AI drafting is a starting point at best. The editorial quality of published journalism must reflect professional journalistic judgment — AI can draft faster but cannot replace the craft of storytelling.
4. Data Journalism and Document Analysis
AI is genuinely useful for analysing large document sets — OIA responses, court documents, company filings, government reports — identifying patterns and anomalies that would take significant manual time to find. AI can help you work through data faster and find the stories within it.
Again: AI findings require verification. Pattern identification by AI is a starting point for investigation, not a publishable conclusion.
5. Social Media and Audience Development
Social media posts, newsletters, article summaries, and platform-specific adaptations of published stories can be produced faster with AI. For newsrooms publishing across multiple platforms, AI can help adapt content for each without the full time cost of manual rewriting.
6. Administrative and Production Work
Pitching emails, grant applications for journalism funding, show notes for podcasts, image caption drafts, and content planning can all be handled with AI support. The administrative overhead of journalism is real — AI reduces it.
Journalism Ethics and AI — The Lines That Matter
Source Protection
This is the highest-stakes AI consideration for journalists. Never enter information that could identify a confidential source into any AI tool. This includes interview notes, document details, and anything that could allow a source to be identified if the AI system were compromised or subject to a legal request.
The New Zealand Press Council and journalistic ethics require robust source protection. AI tools do not carry legal protection comparable to journalist privilege. If in doubt, keep it out of AI systems entirely.
Accuracy and Verification
AI systems hallucinate. They present incorrect information with the same confidence as correct information. For journalism, where accuracy is fundamental to credibility and legal safety, every AI-generated fact must be independently verified before publication. AI can help you find things to check — it cannot do the checking for you.
Transparency with Audiences
The New Zealand media environment is developing norms around disclosure of AI use in journalism. Significant AI-generated content should be disclosed to audiences. The specifics of what constitutes “significant” are still being worked out — but erring toward transparency is consistent with journalism’s accountability obligations.
Copyright and Attribution
AI-generated content used in journalism may have copyright implications depending on the tool and the extent of AI involvement. NZ copyright law is still developing its position on AI-generated content. Be cautious about passing off substantially AI-generated work as original journalism without disclosure.
Getting Started
The most immediately useful application for most journalists: use AI for research synthesis before significant interviews. Take the background documents, previous coverage, and public records relevant to your next story and ask AI to summarise the key points and suggest angles you might explore. Go into the interview better prepared in a fraction of the time.
For newsrooms looking to build AI capability systematically — including tool selection, editorial policy development, and staff training — an AI Assessment provides a structured approach to adoption that protects editorial standards. We work with media organisations and professional communicators across New Zealand.
Frequently Asked Questions
Can AI write news articles?
AI can produce acceptable first drafts for routine, structured news formats (earnings reports, weather, sports scores) and speed up drafting for more complex stories. It cannot do original reporting — interviewing sources, attending events, obtaining documents, or exercising editorial judgment about what matters and why. The craft of journalism requires human journalists; AI can help with the writing production aspects.
What does the NZ Press Council say about AI in journalism?
The New Zealand Press Council has noted AI as an emerging issue for journalistic standards and is monitoring developments. Existing principles — accuracy, fairness, and accountability — apply regardless of production methods. Check the Press Council website for current guidance as this area develops rapidly.
Can I use AI to help with OIA requests for journalism?
AI can help draft OIA request letters, structure document analysis once you receive responses, and help identify what follow-up requests to make. This is a genuinely useful application — particularly for journalists working through large document releases. The interpretation and editorial judgment about what the documents mean remain yours.
Are NZ media organisations using AI?
Yes — NZME, Stuff, RNZ, and other media organisations are all actively exploring and implementing AI tools, with varying degrees of transparency. Individual journalists are also using AI tools independently of organisational programmes. The industry is in active transition; editorial policies are developing in real time. Talk to your editor about your organisation’s current position before adopting tools for published work.
