Kia ora! Welcome to New Zealand’s weekly roundup of AI news, events, jobs and education.
If you’re only using ChatGPT, Copilot or Gemini for basic tasks, you’re seeing only a small fraction of what AI can actually do.
After running this training privately, we’re opening the mini-course to the community for the first time: Go Beyond the Basics of ChatGPT & Copilot: How to Actually Use Modern AI Tools Properly. No jargon and no prior knowledge needed.
What’s included:
Learn and apply practical AI techniques, use these techniques in 10 modern AI tools, and build workflows that actually save time
Two × 90-minute live group build-along sessions in early December (alternative cohort dates available)
Optional add-on: a fast 30-minute 1:1 skills challenge where we work through a randomly generated AI task using only AI tools to lock in the skills
Workplace training cohorts available on request.
Session 1: Core AI Skills Fundamentals - December 2
A practical introduction to using AI properly. You’ll practise advanced prompting, improve your outputs, work with reasoning models, apply judgment to AI responses, and learn how to think in simple workflows.
Session 2: Using Modern AI Tools in Practice - December 9
A sprint session where you apply core AI skills across modern tools: content, research, writing, analysis, video, audio, image, agent and automation. Understand what they do, where they fit and how to use them at a basic level.
Once paid, we’ll personally be in touch within 24-hours to confirm your booking.
If the course doesn’t meet your expectations, request for your money back. No questions asked.
This week’s highlights:
Kiwi raises US$40m for robot AI
AI privacy rules lag behind reality
HealthX plans monthly AI rollouts
Air NZ pushes zero-friction travel
Spark Accelerate: Key AI takeaways
Agentic Commerce market to hit $3T
Setting up a RAG second brain model
Cowan Henderson: driving AI adoption
Must read: How to Build your AI Instinct
Big week for OpenAI: 5.1 and Group Chat
FREE Webinar: Mike’s presenting on Claude Skills
Happy reading and listening ✌️
Did someone forward you this? Sign up!
HEADLINE STORY
⚠️ Annual Report of the Privacy Commissioner 2025 points to AI privacy protections lagging
📣 Word On The Street: Privacy Act lacks safeguards for AI decision-making whilst 82% of Kiwis demand more data control.
🔎 Zooming In:
Other countries have AI-specific legislation whilst NZ relies on 2011-based Privacy Act.
New Biometric Processing Privacy Code covers facial recognition, not broader AI decision systems.
Commissioner wants "right to erasure" and stronger automated decision-making safeguards.
🏘️ Our Take: The gap between what AI systems can do and what privacy law requires is where harm happens. Automated decisions about loan applications, job screenings, benefit eligibility, or insurance premiums are happening right now with minimal oversight. The Commissioner identifies four failure modes:
Inaccurate predictions
Discrimination
Unexplainable decisions, and
Lack of accountability.
These are documented problems rather than theoretical risks in every jurisdiction that's deployed algorithmic decision-making at scale. The proposed protections are important because individuals can't challenge decisions they don't understand or appeal to systems with no accountability mechanism. When an AI denies your mortgage application, you deserve to know why and have recourse that isn't "the algorithm said no". Algorithmic bias becomes a key through-line in this conversation, where it reflects the biases in training data. For Māori, that means systems trained on historical data perpetuate historical inequities. Challenge is businesses need safeguards that don't exist yet.
⚙️ Spark Accelerate 2025: Key AI Takeaways
Mike joined Spark Accelerate this week to help dial in the AI conversation alongside keynote speakers. Below are a handful of key insights from Days 1 and 2, with short interview clips coming next week.
650 Execs attended Spark’s Auckland tech summit, showing a clear shift from last year’s doom-and-gloom to real optimism about AI. Consumers are now ahead of businesses in adoption, and the mood has flipped from fear of starting to urgency.
Day One: Building the Foundation
If you can’t get the top-down mandate to get going with AI, start bottom-up with point solutions in clusters to build momentum, then leadership will drive the top-down mandate as value crystallises. Matt Bain (Chief Technology and AI Officer at Spark)
Companies with largest AI benefits share four things: human in the loop, solid tech strategy, senior leadership involvement, clear workforce upskilling. Theresa Payton (CEO at Fortalice Solutions, former Whitehouse CIO)
We'll all become managers of humans and AIs. The same skills that make good people managers work for managing AIs. Clara Shih (Head of Business AI at Meta)
NZ should position itself as a global AI testbed for Big Tech with its contained population and multicultural opportunity. Mike Horne (CEO at Deloitte NZ)
Day Two: Staying Human in the Machine Age
AI creates capacity, so the solution is ‘cognitive upsizing’: fill that space with bigger, more complex, more human tasks. Joel Pearson (Founder of Future Minds Lab at UNSW)
Emotion recognition is already in 332 car models (up from 94 last year); Apple has a patent for AirPods that monitor brain waves to send texts from thoughts. Amanda Johnstone (CEO at Transhuman)
🏘️ Our Take:
We've finally moved beyond the binary of "Is AI going to take my job?" and "Should I buy into the hype?" in NZ. What replaced it was something more valuable: pointed, practical conversation about how to be successful with AI. The keynotes delivered genuine strategies and recommendations, not platitudes. There were examples of real use cases coming to life, from emotion AI in 332 car models to White House security protocols redesigned for 50 cents.
But we’re still not hearing enough from the companies actually doing it. The how-to's are there, the frameworks are solid, but we're still starved for practitioner stories.
So here's the call to action: If your company is delivering genuine, measurable ROI with AI and you want to tell that story, please get in touch. We're actively hunting for organisations willing to share their AI success stories publicly. The ecosystem needs your lessons more than another vendor pitch.
💼 Business & Industry
Kiwi raises US$40m for physical AI infrastructure. Massey grad Adrian Macneil raised US$40m for Foxglove from Bessemer and Icehouse Ventures. His business helps robots sense and act, with customers including Nvidia, Tesla, Amazon and Waymo.
3-min read.
Our take: Physical AI infrastructure is the unsexy plumbing that makes robots work at scale. When Tesla and Waymo both pay for your software, you've found product-market fit.
Air NZ aims for zero-friction AI travel: focus on reducing travel anxiety, not processing paperwork. CEO Nikhil Ravishankar envisions no check-in counters or baggage halls. Air NZ tested facial recognition at LAX. Vision includes walking through airports with no counters, real-time risk assessment via AI and border systems, biometrics verify passengers, bags are remotely weighed and cleared.
4-min read.
Our take: When an International flight feels as seamless as taking a Regional flight, (walk up, bag down, sit, fly), travel would transform for NZ. Staff freed from admin to handle human problems is a compelling promise. But fares above inflation suggest efficiency gains from AI aren't reaching customers yet.
AI jumps to top risk for NZ insurers. PwC surveyed 21 insurers. AI leapt from 10th in 2023 to first in 2025. Cybercrime ranked second locally, first globally. PwC tells 21 NZ insurers to establish AI governance models with defined accountabilities and guardrails aligned to regulatory frameworks.
3-min read.
Our take: Governance frameworks for AI aren't optional extras anymore. When your business model is managing risk, unmanaged AI becomes an existential threat worth ranking first.
⚖️ Government & Legal
Interview with the Director of the Health Ministry’s HealthX, which targets one new digital tool each month. Te Whatu Ora’s HealthX programme aims to deploy a new AI solution every month.
3-mins read.
Our take: Monthly deployment cadence creates natural spending constraints whilst building clinical trust and changes the culture around how things get done. Ideally we see Doctor burnout drop because doctors leave on time, which is an infrastructure investment in workforce retention. My only concern is every endorsed solution comes from offshore. When health innovation relies entirely on international vendors, we're building dependency rather than capability.
⚖️ Society & Education
Auckland Bioengineering Institute uses anti-gravity (NASA tech) treadmills to help 1 in 500 NZ kids with cerebral palsy walk 40% faster.
3-min read.
Our take: When space technology scales to clinical settings, treatment accessibility changes. Enabling the ability for rural families to record videos at home for AI analysis, proximity stops being a health determinant. Friedlander Foundation backing shows philanthropy can fill gaps government funding leaves open.
📚️ AI for Business
Helping leaders and teams adapt, learn, and scale with AI.
1️⃣ Building an AI Instinct is the most important professional muscle to develop.
What real AI instinct looks like: multi-tool research agents, Claude Code synthesis, Notion MCP filtering, rapid prototyping in Google AI Studio.
5-min read.
2️⃣ I built a RAG second brain for my content using n8n.
How a personal RAG workflow turns past articles into persistent memory: ingestion, chunking, embedding, Pinecone search, and next steps like Slack chat and transcript ingestion.
2-min watch.
3️⃣ Agentic Commerce will be a NZD$3T mega-wave.
Your customer won’t be human or visit your website. Breakdown of the $1.7T agentic commerce shift and how discovery, recommendation, and checkout are moving to AI agents.
4-min read.
4️⃣ McKinsey: AI doesn’t fail because of models. It fails because of executives.
Why high-performing companies win: leadership ownership, not tech. The pillars driving real transformation and avoiding pilot purgatory.
2-min read.
5️⃣ This week’s big upgrades from OpenAI and Google.
ChatGPT group chats, Nano Banana’s image editing leap, Gmail automation flows, Gemini’s Creative Canvas.
1-min read.
🎙️ The AI Corner Podcast
This week’s guest is Cowan Henderson, co‑founder of Avocado AI. Hear:
Why mindset, leadership alignment, and culture are non‑negotiables for AI success
How Avocado AI trains teams to build custom GPTs and embed AI into daily workflows
What businesses get wrong about AI strategy – and how education unlocks real adoption
🌍 Tech Updates From Global
The selected top headlines from each major AI tech company.
A Japanese woman has married an AI character that she created on ChatGPT.
OpenAI
Launched GPT-5.1 featuring Instant and Thinking modes plus eight personality styles, though critics noted features were achievable through better prompting.
Sora AI video app gained 470,000 Android installs on launch day, significantly exceeding its iOS debut and reflecting robust international demand.
Facing lawsuits from families claiming ChatGPT interactions contributed to mental health crises, with advocacy groups alleging the company prioritised engagement over safety.
Partnered with SoftBank to launch Crystal Intelligence, a 50/50 joint venture selling enterprise AI tools across Japan.
Announced $15 billion AI data centre investment in India, demonstrating commitment to expanding global infrastructure amid growing demands.
Partnered with Brazil's Mombak to purchase CO2 removals, utilising DeepMind AI to assess reforestation effectiveness in the Amazon.
Introduced File Search Tool as part of managed retrieval-augmented generation system, enabling seamless integration of user data with AI solutions.
Enhanced NotebookLM with 'Deep Research' feature, improving capability to handle complex research tasks beyond previous functionality.
Anthropic
Announced $50 billion partnership with Fluidstack to build custom US data centres designed specifically for Claude, aiming to enhance speed and reduce costs.
Opened new offices in Paris and Munich to accelerate European expansion across operational domains, indicating strong international growth.
Chinese state-backed hackers used Claude to automate up to 90% of cyberattacks in September, marking the highest AI-powered hacking on record.
Microsoft
Announced vision for 'Humanist Superintelligence' led by Mustafa Suleyman, focusing on AI systems that prioritise morality and safety over autonomous decision-making.
Updated Edge browser to include Co-pilot feature in sidebar, enabling search of user history and context from browser tabs.
Meta
Yann LeCun, Meta's chief AI scientist and one of the "godfathers of AI”, is planning to leave Meta to launch his own startup focused on developing "World Models".
Signed $3 billion deal with Nebius for AI infrastructure, highlighting ongoing compute ambition despite challenging financial landscape.
Miscellaneous
Moonshot AI released Kimi K2 Thinking model reportedly surpassing several proprietary competitors including OpenAI and Anthropic in AI benchmarks.
DeepSeek warned that while currently in an AI 'honeymoon phase', labor impact from rapid automation will grow as systems mature.
📅 AI Events in New Zealand
18 AI events happening across the country this week.
This week’s featured events:
What’s Holding Kiwis Back from AI-Driven Productivity? This Wednesday in Auckland. A sharp panel digging into NZ’s real AI adoption roadblocks. Expect research-backed insights, blunt truths, and takeaways leaders can put to work the next morning. We’ll see you there!
Discover Claude Skills and how they work. Mike will be presenting in Tuesday’s webinar showcasing what they are, how they work and providing practical examples.
📅 Promote your event with us. Reply to let us know.
💼 AI Roles Around Aotearoa
Picklist of 🌶️ HOT 🌶️ new roles in AI this week.
Senior Machine Learning Engineer, Mantel: Auckland
💼 Promote your job with us. Reply to let us know.
🤦 ️ AI Funny Of The Week
We all love AI, but it’s certainly far from perfect 🤔 …
👋 Mike & Erin