Shaping Global AI Safety

The AI Forum New Zealand is proud to have been one of only 80 organisations globally who were invited to contribute to the landmark report, International Scientific Report on the Safety of Advanced AI 2025. Following the publication of the Interim Report, the final version will be presented at the AI Action Summit in Paris, February 2025. 

Given its confidential nature, we established a dedicated Sub-Committee of Executive Council members, led by Glen Willoughby, to provide our response.

This report, chaired by Professor Yoshua Bengio and supported by an Expert Advisory Panel representing 30 countries, the European Union and the United Nations is a critical step toward addressing the safety of advanced AI. 

Our involvement and key highlights

The report acknowledges the dynamic and complex nature of the global AI ecosystem, emphasising the need for international collaboration to balance AI’s immense potential with its associated risks. Key concerns include bias, privacy and the unpredictability of autonomous systems, underscoring the importance of ensuring AI aligns with societal values and ethical responsibilities.

The AI Forum’s response broadly supported the report’s vision, with key points of focus including:

• Indigenous perspectives
  We highlighted the unique challenges faced by New Zealand Māori and other indigenous peoples, including examples of how they are often disproportionately impacted by some of the risks associated with AI. This includes biases in algorithms and limited representation in data sets. We also stressed the importance of culturally relevant resources and interventions to ensure equitable outcomes. This includes the inclusion of AI initiatives that support low-resource languages and cultures, creating pathways toward a more inclusive society.
  
• Data governance
  We emphasised the critical need for robust data governance frameworks, protecting privacy and intellectual property rights, while addressing issues like misinformation through error mitigation strategies.
  
• Technological convergence
  Looking forward, we highlighted the need to explore the opportunities and risks posed by AI’s convergence with other technologies (including robotics), particularly regarding societal and economic impacts. We also highlighted the importance of fostering education and awareness to build societal trust and acceptance of AI technologies.

Special thanks to our Sub-Committee members, whose time and expertise shaped our response:

• Glen Willoughby (Staples VR) 
• Ming Cheuk (ElementX) 
• Prof Michael Witbrock (University of Auckland) 
• Anshuman Bannerjee (SparkNZ) 
• Maria Mingallon (MottMacDonald)
• Karaitiana Taiuru (Taiuru & Associates) 
• Frith Tweedie (Simply Privacy) 
• Madeline Newman (AI Forum NZ)

For further insights, please read the interim report, International Scientific Report on the Safety of Advanced AI, published in July 2024.

---

Keep updated with the AI Forum’s news and events here. If you’re not already a member, learn more and join us.

---