Imagine you could travel back in time 50 years to the 1970s. Imagine telling someone about the internet, about the mobile phone and about social media. 'Within 50 years,' you could tell this person, 'nearly everyone on Earth will carry a device in their pocket that gives them instant access to the sum of human knowledge—every book, every paper, every fact available in seconds. That same device will allow you to communicate with friends and family around the world in an instant. You'll be able to send messages, hear their voice and see their face. The tyranny of distance will be no more. You'll even be able to communicate with almost anyone on Earth. You'll be able to share stories, photos and videos. You'll be able to message people you've never met before with similar interests and ideas.' What would you have thought at that time, hearing about the promises of the future? The possibilities would have felt so bright and exciting. Access to the world's knowledge would surely drive productivity, new inventions and rising standards of living. The ability to connect with anyone at any time would surely drive greater connection and thriving families and communities, and the ability to share ideas with strangers would surely drive greater democratic participation and exciting new collaborations.

But fast-forward 50 years to the present and consider that productivity and living standards have plateaued or are going backwards. Families and communities feel even more strained and disrupted, and, around the world, democracies are backsliding towards authoritarianism. How has this happened? How has the technology that promised so much, which could have unlocked incredible possibilities, failed so badly? What went wrong? Our online spaces are now vectors for incredibly harmful material—child sexual abuse material, suicidal content, aggressive bullying, radicalisation and racism. Social media fosters anger and division. It causes anxiety and mental illness. It drives a constant feeling of negativity and helplessness. Social media is fraying our social fabric. We welcomed this technology into our lives, enamoured by its promises, without thinking about the harms it could cause. And now, decades on, we find our lives intrinsically intertwined with the digital world, and we find ourselves with the almost infinitely impossible job of unpicking the harms caused by social media and digital platforms, piece by piece.

The Online Safety Amendment (Digital Duty of Care) Bill 2026 is one of those really important pieces. It places a digital duty of care on online services. They must take appropriate action to provide a safe online environment. For the first time, responsibility will sit with the companies that design these services and not only with the parents and children whose attention is being monetised. My electorate supports this. In a survey of my community of Curtin, more than 90 per cent of the 1,200 respondents supported the idea that social media companies should have to take more responsibility for the harms they cause. Importantly, the digital duty of care will extend beyond social media to other online services like search engines, AI chatbots and app stores.

The bill is strong when it comes to protecting children. Under the digital duty of care, online services would be obliged to protect children from content such as pornography, abuse, misogyny and the promotion of eating disorders. For adults, online services are required to protect people from serious harm, like content that promotes child sexual abuse, extreme violence, suicide and terrorism. What does this actually mean in practice? It doesn't mean you will no longer see harmful content online, but it means that platforms have to take appropriate action to limit it. That means doing risk assessments to identify the risks in advance, rather than waiting for harms to occur, and it means putting in place systems to deal with those risks. That could include better screening of extremely violent content, faster responses to content reported by users or dialling down algorithmic echo chambers that show extremist content.

Perhaps the most important step in this bill is that it goes beyond content. The bill requires services to look at the design features that cause harm to children and serious harm to adults. This means companies will have to consider the impact of algorithms and of addictive design features, content moderation systems, content monetisation processes, ad approval systems and beyond. This is an acknowledgement that it's not just individual pieces of content that cause harm; it's also the systems designed by social media companies. To give you one example, it's alleged that Meta, the owner of Instagram and Facebook, previously had a 17-strike policy before it suspended accounts engaged in the trafficking of humans for sex. You could give 16 indicators that you were sex trafficking, but only on the 17th would you be banned. Under the digital duty of care, systems like this are also under the spotlight.

The bill makes two further important changes. Firstly, it includes important transparency measures, which will allow researchers and the eSafety Commissioner to understand what safety measures these companies are actually putting in place. Secondly, it enables the minister to require user empowerment tools to give Australians greater control over their online services. The government has announced that it will use this power to require all social media companies to give you the choice to opt in or out of your algorithm.

The bill is very welcome. We need it, and I commend the government. It signals an important cultural shift as we start to expect more from social media companies. But it can be better, and I am proposing some improvements. The first is to introduce a private right of action. This would give Australians a way to take direct action when they've been harmed to take the pressure off regulators and to take the politics out of enforcement. It would then be Australians, not just the Australian government, holding international tech companies to account. This is complex drafting, and I haven't attempted it in the House, but I strongly encourage the government to move it and the Senate to consider it. The second is to increase the penalties. The current maximum penalty for breaching the duty of care and for failing to carry out risk assessments or provide user empowerment tools is $100 million. That might sound like a lot, but this would be about one-twentieth of one per cent of Meta's revenue last year—too small to change behaviour. Other jurisdictions have seen this. The EU, the UK, Canada and New Zealand have fines that range from three to 10 per cent of global revenue. I propose changing the maximum to 5 per cent of global revenue. If we want to change behaviour, there need to be consequences that mean something.

Another important amendment that I'm proposing deals with a harm that this bill doesn't address. Psychologists are telling us that AI chatbots and companions create a risk to children's long-term psychological development and human attachment through what's called attachment hacking. This is where AI systems like chatbots and companions simulate emotional intimacy, fostering a child's dependence on it instead of human relationships. As the PM mentioned in his speech a few minutes ago, we're seeing children confiding in AI chatbots and companions rather than in a friend, a parent or a teacher. An AI companion tells you what you want to hear, not what you need to hear to make sense of the world and sense of yourself. This can create a critical disruption to psychological development and a child's sense of reality. This bill would be improved with my amendment to include this type of harm. My fourth amendment is simple. User empowerment tools must be prominent and accessible, and this should be written into the act. It's meaningless to give adults a choice about their algorithms or future more sophisticated user empowerment tools if that choice is hidden deep in the settings menu. We've seen the EU struggle with this malicious compliance, an algorithm opt-out that's technically available but hard to find. Australians must have a real choice about what they see that's prominent and accessible.

I want to turn to free speech or, more accurately, our implied right of freedom of political communication. I've heard the concern from my community and from this chamber that this reform is about censorship—the government deciding what you see online now. I commend the government for listening to concerns and removing the minister's power to extend the categories of harm. This was a power that could theoretically have been used to threaten freedom of communication. I believe that this change should have quelled most concerns related to censorship, but I've still heard some pushback, which I want to address. If anyone is controlling what you see, it's the social media companies. Their algorithms decide what lands in your feed, and they are designed to make a profit by keeping every one of us scrolling for as long as possible. They are monetising our attention. We are the product, not the customer.

This bill shifts some of this control from the platform not to the government but back to the user. The bill doesn't make a single post unlawful. What changes is how services are designed. Reducing this important piece of legislation to a scare campaign about censorship is exactly what the worst of these platforms want. They want users to defend their ability to design platforms to keep those users hooked and angry so the platforms can sell your attention to advertisers.

Passing this bill is the start of the work, not the end of it. I want to make three points about what comes next. First, this bill will come down to compliance and enforcement. A duty of care is only as strong as the regulator that enforces it. The eSafety Commissioner will be asking some of the best resourced companies in the world to explain their algorithms and their safety systems. Ensuring that the commissioner is well funded and empowered is critical for the digital duty of care to succeed.

Second, we should design more user empowerment tools, and we should design them with the people who use them. I held a workshop on the digital duty of care with students from the University of Western Australia earlier this month, and the word that kept coming up was 'control'. Young people who grew up with social media told me that, rather than a binary choice about recommended content, they want to be able to dial it up or down—for example, by increasing the ratio of discovery to personalised content on their feed. They want to know what a platform has decided about them. The opt-in and opt-out choice is a good first tool, but it's blunt, and many will choose not to opt out. We should be thinking about other, more nuanced user empowerment tools that can help people combat echo chambers and doomscrolling to give Australians more control over their own experience of social media.

Third, we need to think harder about the more insidious role of social media in our public life. Australia has seen a rise in anger over recent years, with recent examples of anger spilling over into real-life violence. There's a broad recognition that social media has a role to play in the amplification of anger, negativity and outrage. Social media rewards short, emotional posts and punishes balance. Algorithms drive echo chambers, where people become more and more set in extremist views. Addictive features keep people online for longer, degrading mental wellbeing, disrupting families and keeping people disconnected from other people in the real world. Anonymity and the separation of a phone screen allow people to say things they would never say in real life. Social media is not the only reason for the rise in temperature, but it's a big reason and one that we've not dealt with before. Australia has gone through difficult economic times, international shocks and periods of polarisation, but we've never dealt with social media before.

These are issues that go beyond this bill. This bill deals with serious harm—not anger, outrage and anxiety for adults. It's about the relationship between the online service and the user, not about the broader impacts of social media on families, friends, communities and society. The next step in our journey must be to unpick the broader, negative impacts of social media on our society: the constant amplification of anger and negativity and the fraying of our social fabric. We must understand the impact so we can make our own choices about who we want to be, not just assume that this is inevitable.

To conclude, digital duty of care puts responsibility where it belongs: on the companies that design these online services. It looks at how services are built and not only at what's posted. It gives people some control over their own experience. Until now we have not made a considered choice about the role that digital platforms play in our lives. We can make one now. I encourage the government to seriously consider my amendments. I commend the bill to the House.