Project summary

AI Safety ANZ (AISANZ) is hosting New Zealand's first-ever dedicated AI Safety Conference (Event Link: luma.com/AI_Safety_NZ). This event, organised and run by a BlueDot graduates and members of the EA Community bridges the gap by bringing together individuals academic research, public policy, industry and education who have previously been working in silos. AISANZ is funding the venue, salaries of the organisers and insurance, we are requesting $5,000 USD ($8000 NZD) to provide travel grants for 16x individuals for a "High-Impact List" of 15+ academics, policy leads, and researchers from across New Zealand, that AISANZ has been engaging and vetting over the last 6 months. These individuals have agreed to offer their time for free to attend, but most have requested or applied for a "travel grant" to compensate their flights and accommodation costs to attend.

What are this project's goals? How will you achieve them?

The conference aims to transition New Zealand from an "AI adopter" to an "AI safety contributor" through four specific pillars:

1. Ecosystem Development: Breaking academic silos by facilitating a sustainable network for collaboration across NZ’s academic, public, and commercial sectors.

2. Strategic Education: Equipping attendees with a technical and strategic understanding of AI alignment and x-risk.

3. Research Visibility: Showcasing impacts of NZ domestic projects run by AI Safety concious individuals (e.g Nelson AI Sandbox, Regulate AI Campaign or The AI Assembly)

4. Professional Pathways: Transitioning talent into the field via exposure to fellowships and industry placements e.g TARA, BlueDot, MATS

We will do this over a 2 day conference where each theme as a specific theme.

Day 1: Education and Theory. The first day focuses on establishing a baseline of critical knowledge regarding the current and future landscape of AI risk. Technical lectures, documentary screenings, and lightning talks on the broad landscape of AI safety. Examples of topics include Cybersecurity, Bioterrorism, Model Evaluation, Productivity, Education, Bias Detection, and DEI.

Day 2: Action and Implementation. The second day shifts from theoretical understanding to applied safety measures and project acceleration. We’ll run Deep-dive roundtables and open discussions focused on solving specific safety hurdles within the New Zealand context. As well as running an ‘UNConference’ model in the final afternoon for attendees to ‘fill the gaps’ and have participant led talks, workshops and roundtables to encourage project leadership and high-value knowledge exchange.

What we’re done so far AISANZ are working directly with academics from the University of Canterbury to source speakers and drawing on our established local AI Risk and Ethics meetup of 350+ members built over 25+ events in the last six months to generate a local audience.

How will this funding be used?

The grant will be used exclusively for travel and accommodation subsidies (estimated at $291 USD / $500 NZD per person). This funding will allow the attendance of high-leverage individuals who otherwise cannot attend due to personal financial constraints, lack of ‘conference’ budgets for Masters and PHD students, current government austerity measures (no conference attendance budget) or consultants who wish for some level of reimbursement for their time invested and engagement with academia.

Travel Grant Recipients in order of priority:

• Ben Cravens: MSc student at Victoria University of Wellington; thesis on "Preparing for Advanced AI: Scenarios, Risks, and Policy Responses.” PHD student of Dr Andrew Lensen.

• Hannah Betts: Projects lead at Far AI and Regulate AI NZ; focus on journalism and transparency.

• Cass Maughan: Master’s Student researching China’s foreign policy on AI Governance.

• Graeme Ford: Open Philanthropy Grant recipient and Director of T Bay Labs.

• Kyle Webster: Physical Scientist and previously Principal Investigator at RAND - Focus on AI hardware safety and biosecurity risks.

• Holly Knill: Recently graduated researcher in AI productivity, with over 20 years of corporate experience. AI representative for the newly formed Opportunities Party.

• Kate Pearce: Head of Security at Trade Me; New Zealand’s leading cybersecurity voice.

• Terry Chapman: Managing Director at Axenic, specialising in AI Assurance and Accreditation for the NZ such as ISO42001 compliance. They have spent over 15 years engaging with the NZ Public Sector such as Ministry of Defence.

• Luis Slyfield: Research Engineer for EquiStamp; conducting model evaluation and red teaming.

• Karrina Mountfort: CTO of Webbased AI, AI Accreditation programs and leader of empowering women in community.

• Richard Brudvik-Lindner: Founder of the non-profit Nelson AI Sandbox, mobilising over 40 volunteers and achieved several hundred thousand dollars in funding from the Rata Foundation to run AI Education initiatives.

• Dr. Andeed Ma: Author of AI for Humanity and leader of the Singapore/NZ branches of the AI Collective.

• Gargi Rathi: Research Fellow from AI Safety Camp and IIT Delhi.

• Sheikh Sultan Aadil Huque: Research Associate at NUS on AI and International Law.

• Public Sector Policy Leads: Including Dylan Chambers (Ministry of Transport) and EA member, and Selena Smeaton (MBIE Head of Data Governance) and engaged in development of New Zealand’s Algorithm Charter.
  

Who is on your team? What's your track record on similar projects?

• Michael Kerrison (Director of AISANZ): Successfully secured foundational funding from Coefficient Giving and director of AISANZ.

• Emma Humphrey: Winner of the £100,000 Innovate UK Fairness Innovation Challenge (2023) for research on CV screening equity. Emma is the community engagement lead for New Zealand, Emma has conducted over 60 high-level academic and policy meetings in the last 6 months to prep the network for this event. She has also grown the AISANZ community from 0 to 350+ active members in Christchurch over the last 6 months via meetups/ lectures and virtual events.
  

What are the most likely causes and outcomes if this project fails?

AISANZ is funding the $2,000 venue, $900 insurance costs, $2000 in tickets for key speakers and volunteers - as well as a part time salary for Emma to organise the conference. The conference will happen regardless; however, the "failure" mode of this conference is a lack of geographic and sector diversity for New Zealand’s first Inaugural Conference.

If we do not have travel grants: The conference remains a predominantly "Christchurch-only" academic meetup. The high-value attendees from Wellington/Nelson/Hamilton/Auckland will be absent. AI Safety development in NZ will continue at an organic, siloed pace over 2-3 years.

This grant compresses that timeline by 12-24+ months by creating a national working group in a single weekend. It also allows AISANZ to move faster than membership driven corporate groups such as the AI Forum whose focus is more on business implementation and establish the non-profit as an industry leader in AI Safety. New Zealand has a unique opportunity to be proactive rather than reactive implementer of these technologies an injection of travel grant funding can demonstrate high impact for a local AI Safety Initiatives and help build credibility as an organisation.

Success Metrics: We will measure success of the event by sending out a survey 6-12 months after the inaugural AI Safety Conference to measure the impact and gather case studies. Examples of key impact demonstrate would be

• Two cross-institutional research papers initiated within 6 months of the conference

• Establishment of a new academic working group at the Centre for Artificial Intelligence and Public Policy (CAIPP) (Otago University) or Centre for Data Science and Artificial Intelligence (Victoria University)

• One formal briefing document delivered to major body such as Nz Ministry Business Innovation and Employment regarding domestic AI risk.

• Transitioning Masters and PHD academics into AI Safety industry roles within the Asia Oceanic Region

• Attendees have taken an Bluedot or TARA Course

• Attendees have secured funding in AI Safety within 6-12 months of the event

• Return attendance to follow up conferences AISANZ intends to host for late 2026/early 2027 in other areas of New Zealand.

Why considering investing in New Zealand AI Safety Development? AI innovation and implementation in New Zealand is largely nascent compared to major centres such as the USA, France, or the UK. New Zealand does not have the data centre infrastructure to compete in foundational AI development — and indeed, New Zealand's own national AI Strategy (2025) acknowledges that NZ's competitive advantage lies "not in competing with Google, OpenAI, or other foundational model developers, but in becoming sophisticated adopters."

The country's first major AI-focused data centre — a US$2 billion hyperscale GPU campus in Southland by Datagrid NZ — only received resource consent in early 2026, with operations not expected until 2028.

What New Zealand does have is a unique ability to become the smartest adopter of high-risk technologies — and to serve as a playbook and active contributor to international frameworks for their safe implementation.

• Despite its small size, New Zealand already leads many larger economies in AI adoption: Microsoft's 2025 AI Diffusion Report placed NZ at ~40% AI adoption — ahead of the USA (28%), Australia (37%), and the UK (39%).

• NZ is a small, educated, English-speaking population and flexible regulatory environment make New Zealand an ideal sandbox for the rollout of new technologies — including EFTPOS at national scale, contactless payments, and driverless shuttles. New Zealand has also pioneered governance in targeted areas, including delivering the world's first Algorithm Charter to address bias and inequity in public sector AI tools.

• New Zealand's Mixed Member Proportional (MMP) political system, with three-year parliamentary terms and accessible sitting members across 120 seats, enables faster law and policy implementation than most democratic nations. A key example is the ban on military-style semi-automatics and assault rifles passed in under 30 days following the Christchurch attacks in 2019.

• NZ is also uniquely positioned as a potential diplomatic intermediary between China and the West: it holds the first free trade agreement between China and a developed nation, was the first developed country to back China's Belt and Road Initiative, and maintains Five Eyes security commitments simultaneously — a rare dual credibility. As the dominant regional power in the Pacific, NZ holds unique sway over Pacific nations increasingly courted by China.

• Combined with its status as a non-nuclear nation under the Nuclear Free Zone, Disarmament and Arms Control Act 1987 and its ratification of the UN Treaty on the Prohibition of Nuclear Weapons, New Zealand has demonstrated a decades-long precedent for building international consensus on the prohibition of dangerous technologies — a model it is well-placed to extend toward an Anti-AI Proliferation Treaty.

How much money have you raised in the last 12 months, and from where?

• Michael Kerrison, AISANZ Coefficient Giving: Funded salaries and organisational overhead over 100K for 1.5 years.

• Industry Ticket Sales: We are charging $200 NZD for industry tickets specifically to fund one academic attendee per ticket sold - since launching over the last month, we’ve secured 13 attendees and $1970 to cover initial catering costs.

• Internal Funding for this InAugural AI Safety Conference: Michael Kerrison/AISANZ is covering the up-front venue and operational costs ($4,900 NZD)