Welcome and Opening Logistics

Liam Carroll

Good morning, everybody, and welcome to the 2026 Australian AI Safety Forum. My name is Liam Carroll. I am an AI safety researcher at the Gradient Institute, and I will be one of the MCs for this event over the next couple of days. And it's a really great pleasure to welcome you all here today. It's pretty amazing to see so many people here for this event over the next couple of days, because we ran the first version of this AI Safety Forum at the end of 2024. That event had about 100 people or so.

And as you can see, this event is much bigger than 100 people or so. I'm reliably told we have close to 300 in the room, maybe a little bit more, and I think maybe 60 or so on the waitlist that we simply could not fit into the venue as well. So it's pretty awesome to see so much growth in just 18 months. But I think it also shows how important this topic is and how wide reaching it is as well. I would like to start today by acknowledging the traditional custodians of the land on which we are gathered.

The Gadigal people of the Eora Nation here, as well as the lands on which our dialling in speakers are presenting from and pay our respects to their elders, past and present. For over 60,000 years. First Nations people have been custodians of this land, caring for it and their communities. And as we engage in discussions during this forum, let us reflect on their enduring legacy of stewardship and think about how we can show similar concern for future generations as we face today's challenges. I would also like to acknowledge that it is NAIDOC Week this week, and I would like to encourage everybody to look into the various events happening around the country to celebrate the week.

So why are we here? First thing to say is that the most recent version of this International AI Safety Report, chaired by Yoshua Bengio and commissioned by the UK government, was released in February 2026. And it's important to say this upfront because this report provides the core framing for this forum. Every talk, panel, or discussion that you see over the next couple of days should relate to this report and its framing of AI safety in some way. And the report has some very clear messages that inform why we are gathered here today.

And the first is no secret at all. AI is moving incredibly fast. This was the exact same bullet point that I had on the first slide for the 2024 forum as well. But 18 months on, it feels like this is true this time. In fact, last time that we gathered for the forum, things like reasoning models and AI agents such as Claude Code were barely in their infancy at that time. And now, for a lot of us, it's probably hard to imagine working without these kinds of models.

So sometimes in the AI world, weeks really do feel like years and years feel like decades, such is the pace of this industry. The report also makes it very clear that the capabilities of these frontier AI systems are increasing rapidly, and they are also diffusing throughout society and throughout the economy very rapidly. And this in turn means that there are also a number of risks that are opened up because of this. The report outlines things like threats to cyber security, labor market impacts, even potential scenarios of loss of control.

And these are all going to be things that are discussed at this forum over the next couple of days. The thing is, the challenges that are outlined in this report are big and they affect all of us in different ways. And to address them, we think that it is very important that this forum tries to foster a coordinated, fast moving response that crosses over disciplinary lines. AI safety is not just a technical problem to be solved with engineering. The many and varied problems require perspectives from lots of different sectors and fields.

But the problem is in Australia, this expertise is quite dispersed across different communities, different disciplines, different institutions. And so with that in mind, the aim of this forum is to turn Australia's dispersed AI safety expertise into a community that can act together on these very hard problems. These challenges are big, they're serious and they are real. But only by coming together as a community can we hope to address them. And I said it at the last forum, and I will say it again here. We do have agency here.

And at this forum we want to exercise it. To do that, we want to encourage everyone here to present and engage with the foundational and frontier knowledge about technical AI safety and AI governance grounded in the International AI Safety Report. We want to encourage everyone to connect with people and missions across research, government, industry and civil society to empower action by the Australian AI safety community. And on this point, I do want to say, last time I felt like I nearly got to talk to pretty much every one of the 100 or so participants, and I know that that's going to be a lot harder at this event with 300 people.

But we really do hope to foster this spirit of conversation and feeling like you can walk up to anybody and have a great conversation with them. And then finally, there is a lot of uncertainty in this field. And to navigate this uncertainty, we really want to encourage everyone to approach this event with a sense of epistemic humility, a sense of curiosity, and a willingness to change your mind. Engage with open minds and open hearts, and please listen charitably. So to turn this into action immediately, I would like to ask everybody to turn to somebody near them.

Hopefully somebody that you don't know, preferably, and ask them what brought you to this forum today. And if one person can speak for a minute and then the next person speaks for a minute, and then we'll move on to logistics from there. I would like to say that the Australian AI Safety Forum is an independent initiative established in 2024 as a public forum for engaging with the challenges of AI safety. This forum has been made possible by the generous sponsorship by the Australian Government's Department of Industry, Science and Resources.

It has also been made possible by the organising committee that you can see here, and is also listed on the website, and you will hear from many of these organising committee members over the next couple of days as well. And I would also like to give a special shout out to the event manager, who is behind all of the various logistics and the amazing website that you see here, Luke Freeman from Good Ancestors. So with that. I would like to invite Andrew Charlton, who I think is hanging around at the back there.

Awesome. So I would like to warmly welcome Andrew Charlton, the Assistant Minister for Science, Technology and the Digital Economy, to give our opening keynote of the forum. Please welcome Andrew.