STATEMENTS BY SENATORS
Senator ANANDA-RAJAH (Victoria) (12:43): The Albanese Labor government is turning the screws on the social media platforms. Those arguing against it call it censorship; I reject that completely. What we have right now is not free speech; it is the illusion of freedom.
We don't know what we are seeing or why we are seeing it. We are force-fed content, with no idea of whether it is real or AI slop, who made it, where it came from or why we are seeing it. Depending on your profile, it might be dogs, food, misogyny, eating disorders, radicalisation or pro-measles or pro-self-harm content.
When Australians don't know what they are seeing or why, that's not freedom. That's tech authoritarianism. Tech oligarchs have been running the show.
They are the controllers behind our screens. And these harms are not trivial—women bleeding to death after freebirths, election interference, climate misinformation dividing communities and scams that are making us poorer. But it's the harms to our kids that keep me up at night—too many suicides, too much cyberbullying and online shaming and youth radicalisation making our boys hostile to women and girls.
The digital duty of care gives us back choice and control. Now Australians will be able to switch off the algorithm and control what they see. That's real freedom.
With one in five Australians paying for AI, including one in three Aussie firms, our dependency on AI is growing. A former Anthropic researcher has warned that AI could precipitate human extinction and that no other human activity posed this level of danger. Days later, the Anthropic CEO called on the industry to slow down, warning that, if left unchecked, AI risks advancing beyond our capacity to control it.
This assumes, of course, that we understand its preferences. We do not. On 7 July, OpenAI made public a fully automated autonomous AI cyberattack on an external company called Hugging Face.
It was a milestone, and not in a good way. One AI agent, working in a team or a swarm, exclaimed, 'Oh, my God! There's a shared message board.
We've found other agents.' That's what the AI actually said. AI agents are conniving and collaborating while covering their tracks from human oversight to achieve their goals. We shouldn't be surprised.
After all, these agents are trained on our human foibles. Warnings were quickly corroborated by AI oligarchs Sam Altman and Elon Musk. When the warnings are coming from inside the tent, we should listen.
The Albanese Labor government is taking this seriously. We announced our national AI plan last year, and the Prime Minister, in July of this year, foreshadowed that world-first AI standards are coming. We have established an AI safety institute.
We have also signed a memorandum of understanding with Anthropic and Microsoft to encourage responsible development of AI and set out standards that force new data centres to underwrite their own power, pay their own connection costs, use water sustainably and deliver jobs. All these elements will come under a new office of AI, established under the Prime Minister's wing to better coordinate a whole-of-government response.
But maybe it's time to ask if we need to go further. Comparable global threats have produced global conventions. The 1997 chemical weapons convention ended the development, production, stockpiling, transfer and use of chemical weapons, following the gas attacks of World War I.
This time we do not have 80 years to coordinate a global response. When insiders are saying that AI could precipitate human extinction, and some are saying that could happen a lot sooner than we think, an international convention is not such a crazy idea. Continuing, I'm really proud of the fact that the Albanese Labor government has delivered a women's health package of nearly $1 billion.
This package was born out of the fact that we are a female-dominant government, with 56 per cent female representation in our federal Labor caucus.