View the videos of the 2026 Australian AI Safety Forum
On 7–8 July, 329 participants and 79 speakers from research, government, industry and civil society met at the University of Sydney for the 2026 AI Safety Forum (AISF). The event was roughly three times the scale of the inaugural 2024 edition.
If you missed the event or could not make some of the sessions, you can now watch the video recordings on the AISF official website.
Why hold such an event, and why now?
The International AI Safety Report, a comprehensive review of the latest science on the capabilities and risks of general-purpose AI systems, makes clear that addressing AI risk effectively requires the joint effort of many disparate disciplines and expert communities.
It asks researchers, regulators, policymakers and industry to move together, and to move quickly.
Australia is not short of the necessary expertise.
That expertise is, however, distributed across organisations, communities, and disciplines that seldom convene together.
To address this, the Forum was built around three goals:
Knowledge Put the current state of the science before the room, the well-established alongside the genuinely open, with the International AI Safety Report as the grounding reference.
Connection Bring together people across research, government, industry, and civil society to facilitate collaboration and collective action on AI Safety in Australia.
Agency
Meet the uncertainty of the moment with curiosity and epistemic humility. Much of what is uncertain about AI's trajectory is uncertain because it is not yet determined, and that is precisely what leaves room to act.
In essence, the forum exists to turn Australia's dispersed AI safety expertise into a connected community that can engage seriously with the urgent and formidable challenges that AI poses to us.
Assistant Minister Andrew Charlton opened the Forum. Speakers included the leadership of the new Australian AI Safety Institute, international researchers, and Australian experts across various sectors.
We sincerely thank the volunteer organising committee and their organisations: Gradient Institute, RAND, Timaeus, Good Ancestors, the University of Sydney and CSIRO, as well as the Department of Industry, Science and Resources, every speaker, facilitator and participant, and everyone else who worked to contribute to the creation and execution of this event.
We look forward to convening this community again, and to the work that happens in between.
An interest list for the next edition is open at aisafetyforum.au.
