Introduction
This is second major instalment of the International AI safety report since its publication in 2025. It was written by independent experts and aims to provide scientific information to support informed policymaking.
Australia’s representative on the expert advisory panel was Dr Liming Zhu, Research Director, Data61, CSIRO. Australian civil society and industry contributors included the Gradient Institute, Old Ways New, Harmony Intelligence and Good Ancestors Policy.
The International AI safety report aims to synthesise scientific evidence to support informed policymaking. It does not make specific policy recommendations.
The report finds AI capabilities have continued to improve. It also presents increasing evidence that AI systems are creating real-world harms:
- They can be used maliciously (e.g. to facilitate cyberattacks).
- They can malfunction (e.g. provide misleading information).
- They also pose systemic risks (e.g. labour market risks).
The report also provides an analysis on future uncertainty. Between now and 2030, AI could slow or plateau, continue at current rates, or accelerate dramatically. This highlights a further need for developers, researchers and policy makers to adopt appropriate risk management practices to prepare society for the future impacts of AI.