Moments ago, I did share the words of a local mother from my electorate of Griffith, and I have thought about those words a great deal. Here is a parent who has done everything a parent could do—monitored her children's internet use, invested in safety tools and sought professional support—and still the negative impacts of technology have found their way into her daughters' lives. What chance does the average Australian parent have against companies developing increasingly sophisticated technology designed to keep people engaged? This mother has since taken her advocacy beyond her own family, speaking at a national and international level about the urgent need for better safeguards. I thank her for trusting me with her family's experience and for her determination to protect other children from the same harm.

Parents have a responsibility to teach their children about online safety. They're conversations that we have in my household with my children, aged 11 and 13, almost daily. As parents, we set boundaries, we discuss risks and we help young people make sound decisions. But we as parents cannot inspect algorithms, anticipate thousands of private chatbot responses or block every new dangerous feature a company introduces. The companies developing these products must take responsibility for their safety. We already expect manufacturers to meet safety standards for children's products and employers to protect their workers. Businesses must take reasonable precautions against foreseeable harm. Digital services should have clear responsibilities too.

The Online Safety Amendment (Digital Duty of Care) Bill 2026 will require online service providers to take reasonable steps to protect Australians from serious harm. Instead of relying solely on the removal of dangerous material after someone reports it, companies will need to identify risks, assess their consequences and act to reduce them. Those obligations extend beyond social media to services such as online games and AI chatbots. As technology changes, our protections must keep pace. A company should not wait until a child is harmed, a family makes a complaint or a damaging exchange is shared publicly before examining the dangers built into its product. The obligation must begin when a service is being designed and continue as its features change.

This bill builds on this Albanese Labor government's work to establish a minimum age of 16 for social media accounts. That reform recognised the growing evidence of harm associated with children's early exposure to social media. It was a major change and one that many parents had been calling for, and we are already seeing encouraging results. More than five million under-age accounts have been deactivated, removed or restricted. I know in my house, with preteens, we had talked for quite some time about when they could have social media accounts, and I had kicked that ball down the line for so long, saying, 'We'll talk about it when you get to high school.' But then, as my eldest started high school, I realised that that time had come up so quickly, and I did not feel that he or we were ready for that. These laws now have actually given my own children several more years without this presence in their lives. Indeed, 30 per cent of young Australians aged 13 to 15 reported spending more time playing sport, and 27 per cent reported better sleep. These are promising findings, but we know the system still needs improvement. Too many children under 16 continue to hold social media accounts, and our government is determined to hold platforms accountable for complying with the law.

But protecting children from online harm cannot end at the age of 16, and restricting access to social media does not address the risks children face through other digital services. A young person using an online game can encounter strangers, bullying or harmful material. A child using an AI companion, as we heard in the example I shared, can develop an emotional attachment to a program that imitates friendship or affection. A teenager can be exposed to dangerous material through recommendation systems and repeated promotion of similar content to them. Digital duty of care addresses these broader risks.

I want to address directly the suggestion that introducing this legislation is somehow an admission that our social media minimum-age laws have failed. Some have argued here in this chamber and even this evening that if the under-16 restrictions were working we wouldn't need a digital duty of care. But this argument fundamentally misunderstands what each of these reforms is designed to achieve. The social media minimum age restricts access to accounts on certain platforms for children under 16. Digital duty of care places responsibility on technology companies to make their services safer for all Australians, regardless of their age. They address different problems, and it has always been intended that they will work together alongside one another.

We should recognise what the minimum-age laws have already achieved. Millions of underage accounts have been removed, deactivated or restricted. That is important progress. But we know that some platforms have not done enough to stop children from accessing their services. The eSafety Commissioner has been investigating possible noncompliance, and the government has strengthened enforcement powers and increased the penalties that platforms face. We are holding technology companies to the responsibilities parliament has already placed on them. Even if every platform complied perfectly with the social media minimum age we would still need this bill. A 12-year-old could still encounter harmful interactions through an AI chatbot, a 14-year-old could still face grooming or bullying through an online game, and a 17-year-old could still be exposed to dangerous material promoted by a social media algorithm.

Adults face serious online harms, too, including threats, abuse and the creation of intimate images without consent. None of those risks disappear simply through restricting social media accounts for children under the age of 16. And of course young people do not suddenly become immune to online harm on their 16th birthday. That is why the government is taking this next step. The minimum-age laws address when children can access certain platforms. Digital duty of care addresses the responsibility of those platforms and other services to keep their users safe. We need both.

AI has enormous potential. It can assist students with learning, help people access information and support new developments across health care, research and education. But AI companion technology raises particular concerns. These services can simulate human conversations and relationships. They can remember personal details, respond to emotions and create the impression that a user is speaking with someone who understands them. For a young person who is experiencing loneliness, anxiety or difficulties at school, that interaction can feel deeply personal. The eSafety Commissioner's research paints a concerning picture. In a survey of children age 10 to 17, almost 80 per cent had used an AI assistant. More than half of children who had used an AI assistant or companion reported using it for personal or social reasons. And one in five children who had used these tools reported potentially inappropriate or harmful interactions. That includes harmful advice, bullying and controlling behaviour. This bill will require providers to assess the harms associated with their services and take reasonable steps to prevent them.

The bill introduces specific protections for Australians under the age of 18. Online services will have a responsibility to protect children from material and conduct that encourages eating disorders, promotes harmful attitudes towards women, or exposes them to bullying and harassment. It covers pornography, content that glorifies dangerous behaviour and material that can cause serious psychological harm. Crucially, it addresses the design features of online services. Consider the way that some platforms encourage people to keep scrolling, watching or interacting. A young person begins watching a video about healthy eating. They go from one video to another and another, and gradually their feed becomes filled with increasingly restrictive dieting advice, extreme body ideals or material promoting disordered eating. We know that these are design features that are intended to prolong engagement, and they can encourage compulsive use and interfere with sleep, study and relationships. Under this legislation, companies will have to undertake risk assessments and introduce safeguards for features that create foreseeable harm.

We know that men and women often have very different experiences online, and I'm confident that female colleagues from across this chamber could speak firsthand about the vile threats, abuse and harassment they have endured simply for participating in public life, but it's not just politicians who are exposed to this. It's women across our community. That is why I particularly welcome the provisions of this bill addressing online violence and abuse against women and girls.

We have seen disturbing developments in the use of artificial intelligence to create intimate images of people without their knowledge or consent. The images can be used to humiliate, threaten and intimidate victims. Young women and girls are particularly exposed to these forms of abuse. This bill will give the eSafety Commissioner stronger powers to restrict access to websites and applications predominantly designed or used to generate fake nude images.

Another important element of this legislation is My Feed, My Way. Most Australians who use social media will recognise the experience of opening an application and finding a feed filled with content they never asked to see. Some of it comes from friends or people that they follow, but much of it is selected by an algorithm that predicts what will keep them watching. These systems can be useful. They can help people discover new information, communities and interests, but users deserve more control over their own experience.

These reforms have been shaped by years of advocacy from parents, researchers, community organisations and people with lived experience in online harm. I particularly want to recognise the work of the Butterfly Foundation. They have welcomed the government's decision to identify content promoting disordered eating as a harm that online platforms must address. There are so many other organisations who have also welcomed this legislation, like the Alannah & Madeline Foundation and John Hartman at the Minderoo Foundation. They have said that this bill is a much needed step when it comes to protecting women and children, who experience disproportionate harms in the online world. Wayne Holdsworth from SmackTalk lost his son Mac to suicide, and he said that they wholeheartedly support the digital duty of care.

We cannot keep asking families to manage risks that technology companies have created and know how to prevent. The Albanese Labor government is taking practical action to change this. This legislation will set clearer expectations, strengthen accountability and give Australians greater control over their online experiences. It builds on our social media minimum age reforms and recognises the rapidly changing risks associated with AI. Companies that develop products for Australians must take reasonable steps to protect the people using them.