MOTIONS

Senator DAVID POCOCK (Australian Capital Territory—Independent ACT Whip) (10:17): I thank Senator Shoebridge for bringing forward this motion. This is something that should have been a concern for the government and the parliament well before this latest warning from current and former Anthropic employees. We've been warned for some time now about the risks of artificial intelligence, the rate of progress and the AI arms race that is currently underway.

Even if, as some people in the government have suggested, this is just AI companies trying to pump up their valuation before a potential IPO, there are many who aren't in these companies who've been warning about similar things for much longer. If any of us here today were offered a ride in a car and, before we jumped in, the driver said, 'Listen, I don't know, but there's maybe a 10 per cent chance that you'll die getting to your destination,' I don't think any of us would be getting in that car.

So what are we doing here? How is this not a matter of urgency for the government? How do we have an AI safety institute that's simply in a department rather than being a statutory authority, like they have in other countries, that's well resourced, like they have in other countries, and that can be looking at what is unfolding and providing independent, robust advice to the government and to the parliament.

I'm concerned that we are faced with a challenge that our government doesn't seem capable of responding to in the time that is required. We've seen nothing from the government other than meetings and photo opportunities with CEOs of AI companies and US government officials. These AI companies are saying: 'Please regulate us.

We're in this position where our motive is profit, yet we're obviously humans and we can see the potential risk to ourselves and the planet. Please regulate us.' But we've got a president of the US, our closest ally, who says: 'No. We want to win.

We want to win this.' It's very unclear what he wants to win, but he just wants to win. I think there is a huge, huge problem for us, as a country that seems to just do whatever the US wants rather than actually standing up, as we've seen Mark Carney and other world leaders doing, and forging our own path. I think that is why it is so urgent that we debate this motion today: to send a very clear message to the government that, as a parliament, as a Senate, we expect them to be working with the Canadians and to be working with other middle powers to rein in frontier AI model development in the US and in China.

We can't just be saying: 'Well, it's too hard. It's too hard. We're a middle power.

The best we can do is allow these big hyperscalers to come here, use our land, use our energy, use our water, train their models and not ask the question: for what?' For what? Why are we rushing into this? It seems like there's no clear explanation from the government apart from, 'We don't want to be left behind.' Yet we haven't clearly defined what we want from this technology.

As Geoffrey Hinton, who's been talking about this for many years now, was saying, and as I read this morning, there is huge potential for this technology. We're already seeing it transform the way we do medical research, plant research. There's a massive upside.

But if there is the downside that we're being warned of—and, sure, sometimes it's hard to tell the difference between genuine warnings and tech bro CEOs just trying to pump up their companies' valuations—we should be listening to experts. We should be taking a precautionary approach, which is, sadly, something that Australian governments haven't seemed to have understood for a very long time.

It seems like we're going to have to wait until we're in another awful crisis to get a government to act. Hopefully, Australians will look back and say: 'You had the crossbench urging you. You had experts urging you.' Now is the time to get on the front foot with this, and I welcome this urgency motion today.