Consider the influence that social media companies have over you. Consider the way they shape the news you read, the jokes you listen to and the people with whom you interact. When your feed was chronological, simple and all about the people you actually knew feels like a distant memory. Most of us—at least, those my age—joined social media to connect with friends and family. Today, the algorithm can be fun, interesting and even informative, but it can also be divisive, isolating and radicalising. It exercises enormous power over us, and we barely understand how it works.

Governments have allowed tech companies to cultivate this power with minimal oversight, and for too long a handful of tech billionaires have written their own rules. They've had years to fix the problems themselves, and they have chosen not to. These multinational companies have discovered that the most effective way to increase their profits is to hook us—and the best way to hook us is to inflame us, to drive us further apart from each other into echo chambers and digital worlds that are often dark, dangerous and divisive. In just 13 minutes, an account mimicking a 16- to 18-year-old boy will be fed misogynistic content. Too many children in my electorate are experiencing these harms, and parents write to me frequently to express their concerns. They feel powerless to protect their kids from these formless, invisible algorithms that manipulate their kids' minds and distort their development.

Social media is the Wild West, and the Online Safety Amendment (Digital Duty of Care) Bill 2026 tries to bring some law and order into the digital world. It's time for government to crack down on the tech-bro cowboys and govern to protect the people who put them in this place. At its core, this bill puts a legal duty on online services to take appropriate action to keep Australians safe. That covers social media, websites, messaging and online games. For the first time, companies will be required to protect us from serious harmful materials and conduct resulting from the products they sell. These serious harms include child sexual abuse material, grooming and image based abuse. In addition, companies must protect children from specific harmful material, including content promoting eating disorders, self-harm, misogyny and dangerous stunts.

Tech companies will need to give users new tools that increase an individual's ability to control what they see, rather than be subject to the invisible whims of the algorithm. At least once a year, they will have to assess the risks their own products create and act on them. The cost of taking action is no excuse. There'll be fines for breaching the duty of up to $109 million. These could and should be higher, and that's why I'll be proposing a sensible amendment to calculate penalties as a percentage of revenue. This would make our legislation consistent with similar legislation in other jurisdictions which places higher penalties on companies, recognising the enormous quantum of profits that these companies enjoy.

The bill also gives the eSafety Commissioner new powers to take down nudify apps. It halves the time platforms have to act on cyberbullying and abuse complaints before the eSafety Commissioner can step in, from 48 hours to 24 hours. It starts to prise open the black box of platforms so that independent researchers can finally look into how those algorithms actually work.

Opponents of this bill claim that it's about censorship, that government wants to control what you see and read. The obvious response to that objection is: if there's someone controlling what you see, it's the social media companies that develop the algorithms, with no scrutiny or guardrails—and they do so entirely in pursuit of profit. This bill is about giving you more control to determine what you watch, what you listen to, what you read and, therefore, how you think and feel.

Let's just take a step back and remember what it is that's being restricted here—child sex abuse material, terrorism, nonconsensual fake nudes, rape threats. Which of these do you want to allow unrestricted? For children, preventing their exposure to pornography, to material that encourages disordered eating or to attitudes hostile towards women seems, to me, eminently sensible and reasonable. I welcome the government's decision to drop the power for a minister to add new categories of harm without first coming back to the parliament. Limiting this ministerial discretion, I believe, is an important step and should put to rest the claims that this bill is about censorship.

The bill could be stronger. I'm going to support the amendments put by my colleagues from the crossbench. It should ensure that opt-in to algorithms is the default, and that's what the campaign led by Chanel Contos of Fix Our Feeds is calling for. It would mean that you have to intentionally choose to use particular algorithms rather than being nudged into them by the tech companies. The bill could include amendments to protect children from AI companion chatbots, including attachment hacking where tech companies manipulate people into emotional connections with AI. I applaud, in particular, the member for Curtin for her leadership on this issue, showing what constructive, courageous leadership from the crossbench can look like.

More broadly, this bill doesn't address every harm caused by social media and other online services. We still have so much more to understand about how these services and the companies operating them drive radicalisation and polarisation. The bill's own explanatory memorandum points to research showing social media can accelerate young people's paths into violent extremism. We see the same forces pulling our politics further apart, making it harder to talk with one another, disagree respectfully, trust each other and find common ground. If we want safer and more connected communities, it's time to put the platforms in their place and the people in control. I commend the bill with the amendments brought by the crossbench to the House.