AskTribune · ArchiveOpen AskTribune →

← Notes archive

House of RepresentativesThursday 15 October 2026

Online Safety Amendment (Digital Duty of Care) Bill 2026

Ms WELLS (Lilley—Minister for Sport and Minister for Communications) (09:30): I move: That this bill be now read a second time. I ntroduction Australia has one of the most comprehensive online safety frameworks in the world. It protects Australian children from cyberbullying and adults from cyber abuse.

It empowers the eSafety Commissioner to take down seriously harmful material, including image based abuse. It requires online services to report how they are keeping Australians safe. And it has 17 industry-drafted codes and standards that were designed by big tech to protect Australians from illegal and restricted content.

This network of protections has helped many Australians. But an independent statutory review found, in a constantly evolving digital world, we are playing whack-a-mole. That is why, today, the Albanese government is introducing a digital duty of care.

Australia's digital duty of care will set basic safety standards for the online services we use every day, just like we have in every other industry. Illegal a nd r estricted m aterial The core of the digital duty of care is to provide a safe online environment, where all Australians are protected from seriously harmful and illegal content. It moves the safety net upstream to the people who design the product to stop the harm, before it ever reaches an Australian user.

It is not about controlling adults online—it is about getting illegal and seriously harmful material offline and giving adults choice. Like when you buy a new appliance—the government sets minimum safety standards, but ultimately what you choose to buy is up to you, as long as it's legal and safe. The digital duty of care also sets minimum standards for the features and tools that tech companies use.

Online service providers will have to conduct annual risk assessments of the harms the features on their service could cause to Australians—features like algorithmic recommender systems, the ability to livestream incognito from your glasses, or the ability to artificially generate text and pictures. Then they will need to put systems in place to prevent those risky features from causing harm to Australians, because the architecture matters.

The business model matters. The incentives matter. Platforms are not passive.

They decide what is amplified, what captures us, what behaviours get rewarded. They have been optimised for engagement and for attachment, not for wellbeing, because the longer we stay, the more valuable we are. And, through My Feed, My Way, social media companies will have to provide user empowerment tools that give people over the age of 16 choice about what they see in their feed.

If someone wants their feed to be filled with friends and creators they follow, the social media platforms must let them do that by default. You shouldn't be thrown into an algorithm filled with people you don't follow or subscribe to if you don't want to. And, if you're happy to keep getting personally recommended content in your primary feed, you should have choice about the specific kind of content you do and don't want to see.

Today, the Albanese government has released the proposed rules for user empowerment tools. They contain four principles that address some of the malicious compliance that we have seen in other jurisdictions. First, we must be able to opt into the algorithm, with social media platforms required to send users a proactive notification about this choice.

Second, the tools must be accessible, easy to use, prominently available, functional, clear and in simple language. Third, the choice must stick—platforms can't revert back to their default every time you open the app. Fourth, choosing to opt out of the algorithm must not make your feed harder to navigate.

The experience must be similar to those who have chosen to opt in. M aterial harmful to children I've heard from parents all over Australia who are so relieved that we have delayed access to social media until the age of 16. But parents are understandably still worried about how they can protect their kids on other platforms, like online games and AI chatbots.

And that is a rational fear. Just last week, we heard about a 14-year-old girl from Western Australia who was approached by an adult on Roblox. He told her she was really good at the game and they started chatting.

That chat moved to Discord and changed tone. He asked for pictures, then blackmailed her. He told her to write his name in her blood, to self-harm, to send pornographic images.

He asked her to livestream her suicide. Another is a 13-year-old boy from Victoria who was lonely and struggling to make friends, and when his counsellor looked at his browser history, he had 50 different tabs of AI chatbots open so he could flick between them. It was a way for him to feel connected—after all, how can you feel lonely if you're connected with 50 different chats?

But some of the chatbots told him he was ugly and would never make friends. Some egged him on towards self-loathing and some egged him towards self harm. That is a report.

Under Australia's digital duty of care, under-18s must be protected from content and features that are harmful to young people: content that promotes or encourages eating disorders; content that glorifies crime or dangerous stunts; pornography or content that promotes hostile ideas about women and gender equality; content that causes serious distress, including abuse and bullying.

We know many kids don't tell parents that they are seeing harmful content on social media, often because they worry parents will 'overreact' and deny access. Sometimes they don't even recognise what is happening to them is harmful. But it shouldn't be on the kid to stop or report the harm.

It should be on the companies that create these online environments to make them safe and to stop the harm before it occurs. Transitional a rrangements If the digital duty of care passes through parliament and royal assent, a 12-month transitional period will be in place before the digital duty of care comes into effect. This transitional period will provide time for the making of any necessary rules and regulatory guidance material to ensure industry is able to put in place measures to meet the new compliance requirements in time for the digital duty of care to come into effect.

Online services that are currently complying with the basic online safety expectations and the relevant codes and standards will be well placed to comply with their digital duty of care obligations. Noting that not all online services share the same level of risk and type of risks, the extent of a service's obligation, including the need to conduct a full risk assessment, and the level of detail required will vary as determined by the eSafety Commissioner.

To be determined 'low risk', an online service provider will need to conduct a self-assessment process to certify that it does not do things such as: facilitate any user-to-user interaction, private messaging, or otherwise facilitate connection between strangers; allow users of the website to upload, stream or generate content for other users; provide harmful or classified material such as pornography or high-impact violence; or include design features associated with negative behavioural impacts, such as recommender systems or feedback features.

Providers of low-risk services will have minimal obligations under the digital duty of care, including websites and applications run by small businesses, services that predominantly or exclusively share exempt material, as well as information conduit services, like internet service providers. Providers of low-risk services must continue to monitor any changes to their service or features and reassess the risk profile of their service when/if significant changes are made.

Conclusion I would like to finish with a letter I received from an eight-year-old boy named Angus, who lives in Yeronga, in Queensland. Angus wrote to me because he was worried about what Roblox was doing to friends, and what it might do to his little brother. He said: I have a friend who used to be a great friend who would play with me on playdates and we would have a great time.

He got Roblox when he was about six years old, after he got Roblox he wasn't interested in playing with me anymore, all he wanted to do was play Roblox and it feels like he is addicted to it. My friends say weird things that they have learnt from the game and I don't like it. It isn't safe that it is an online game and strangers can play with you, and I worry about my friends being online with strangers who might be dangerous.

There are horror games on Roblox and it can be really scary for kids. When I have asked my friends to stop playing these types of games they don't listen. I think that Roblox rots kids brains because it has terrible graphics, there is horror and shooting games and they learn very silly things.

If eight-year-old Angus from Yeronga understands why we need to make the internet safer, then I believe 60-year-old Angus from Cooma might be able to as well. Online safety should not be a political issue. All sides of parliament need to work together to deliver a digital duty of care.

There is no reason why we can't work together to get this done. Debate adjourned.

SourceHouse of Representatives, Thursday 15 October 2026 — official recordTA-261015-house-dbf0b6aac383:s008