Online Safety Amendment (Strengthening Enforcement for the Social Media Minimum Age) Bill 2026

Senator PAYMAN (Western Australia—Australia's Voice Whip) (18:41): The social media ban is the quintessential Albanese government policy. There was no evidence to support its implementation, but that doesn't matter when you're 'leading the world', as we keep hearing so often. It was exposed to a vigorous five-day committee inquiry before being rushed through the parliament.

It was not a reform designed to support Australia's young people but a scheme to win votes by exploiting the justified fears Australian parents have about social media. As has long been the government's messaging, anyone who criticises the policy is in the corner of the big tech companies attempting to stop the government from protecting Australian children. At estimates in October I was told by the minister at the table in a very particularly undignified outburst, 'You support the platforms over parents.' It sounds catchy, but it's not true.

The ban currently allows Aussie kids to access violent, sexual and extremist content without regulation on smaller websites like the notorious 4chan, which eSafety described in an answer to a question on notice as 'one of the most frequently reported suicide related websites to eSafety and has been implicated in deaths by suicide of Australians'. Isn't that supporting platforms over parents?

The ban allows kids to go on chat services like Discord and Roblox where predators lay in wait. Child sexual abuse advocacy organisation Bravehearts cites research on its website which found that 17 per cent of Australian children and young people had been subject to online sexual solicitation by an adult. For girls, this was as high as 26 per cent.

The median age at which the solicitation occurred was 14 years old. Following reports in the Guardian, Roblox has begun to implement age verification, but it still isn't part of the ban. Isn't that supporting platforms over parents?

One of the other big concerns with social media is the ability it has to hook people in. Algorithms designed to keep you scrolling and change your brain chemistry are everywhere. The ban prevents young people from having a YouTube account.

It does not prevent them from watching shorts for hours on end in a logged-out state. That's hour after hour of brain rot content, which is no different to before the ban. Isn't that supporting platforms over parents?

It is the natural instinct of the Albanese government to rush through big reforms to avoid scrutiny. The bill spent just eight days before the parliament. It was introduced in the Senate on 27 November 2024 and passed on the 28th.

As I mentioned, the inquiry into the social media ban lasted just five days, and submissions were open for a mere 24 hours. What were the key recommendations of that rushed inquiry? No. 1, the recommendation that the committee believed was most important, was 'that the Australian government legislate a digital duty of care to place a legal obligation on digital platforms to take proactive steps to protect their users'.

It has been nearly two years since that recommendation was made, and we're still waiting for that digital duty of care. I was told in estimates last year that the Minister for Communications was excited about the reform, but officials from the department have said it may not come into effect until 2028. That's another two years that young people will be waiting for this digital duty of care.

While I have held out hope that the government will finally make real change in the online space, I am concerned that what will be brought before the parliament is, among other things, overly reliant on delegating legislative power to the executive. The third recommendation of the committee report is 'that the Minister for Communications provide a progress report to the parliament on the age-assurance trial by no later than 30 September 2025'.

I had to force the government to produce that report, using the Senate's power to order the production of documents. Government senators opposed the production of the report. According to the Parliament House website, the document was ordered to be produced in August 2025 and wasn't provided until March of 2026.

We know why the government didn't want us to see this report on age-assurance technologies. An article in ASPI's the Strategist explains: Facial recognition tools, meant to be a privacy-friendly verification option, have proven spectacularly unreliable. Eleven-year-olds were being identified as 30.

Sixteen-year-olds legitimately old enough for access were getting locked out. Tech-savvy teens discovered that drawing on fake facial hair with makeup, and in one case recruiting a pet dog, was enough to fool the algorithms. Recommendation 7 of the committee inquiry was for the statutory review of the social media ban to be brought forward by 18 months.

The review timeline was brought forward—but only by 12 months. The implementation of the social media ban has also been one of high secrecy. Many senators in this place, including me, sought details of the government's claim that 4.7 million accounts had been deactivated after the Australian Financial Review reported: … sources familiar with the figures who requested anonymity to speak freely said the 4.7 million number is not a fair representation of the number of active social media accounts belonging to users under the age of 16.

The report went on to say: The figure does not represent 4.7 million individual users because it also includes inactive and duplicate accounts, as well as some that had already been deleted but still exist in the platform's backend. How did the government respond to this? 'Oh, it's not in the public interest to provide that detail.' They wouldn't provide platform-by-platform numbers—how many accounts existed before the ban, how platforms were administrating age verification.

The only number available was 4.7 million, and it was being used in every press conference and at every opportunity to showcase how this world-leading policy is being effective. So show us the numbers. Show us the breakdown.

In August last year, the Senate asked the government to produce documents about the implementation of the ban. Some documents were produced while others were withheld because their publication would 'damage commercial interests'. 'What damage could the disclosure cause?' the Senate asked. 'We're not telling,' the government replied. In November, the Senate asked about the expectations the government had of a constitutional challenge to the social media ban.

Again, the government defied the Senate, despite there being no precedent for withholding such information. They would not even confirm through estimates questions on notice whether they had sought advice as to the constitutionality of the ban. So after all the rushing and all the secrecy, surely the ban has done the one thing it set out to do.

Surely, it has taken kids off of social media, right? In April, eSafety found that 70 per cent of parents of children subject to the ban reported continued access to social media. In June, a study led by the University of Newcastle found 85 per cent of participants aged under 16 were using age-restricted social media platforms.

Thus began the campaign for this, the social media ban 2.0, which is over before it began. This is not something you can just fix up. This is not something you can just keep trying until you get it right.

If there ever was a chance of this working, which I doubt, it was on 10 December 2025. Noncompliance is now embedded, as eSafety warned the government may happen. No matter how much tinkering around the edges occurs, kids will find ways to continue hanging out with their friends online.

Increasing existing penalties is a classic Albanese government strategy to look like they're doing something when public confidence in the government's ability to manage a particular issue is floundering. The parliament recently passed a bill to increase penalties for illegal tobacco offences, which follows more than $300 million of investment in compliance in the last two budgets as the black market gets bigger, not smaller.

Amidst the fuel crisis caused by the war in Iran, allegations of price gouging led to increased penalties for breaches of the Competition and Consumer Act. During the two-day sitting in January, laws were passed to increase penalties for existing hate speech offences. Other elements of that bill are likely to be struck down in a High Court challenge that is being heard today.

The doubling of these penalties is a message to the Australian people that the government actually don't know how to fix the social media ban, but they want to look like they do. The oft-cited but unused $49 million penalty provisions for failing to take reasonable steps to prevent kids under 16 from holding accounts will be doubled to $99 million. Why not a billion dollars?

Why not a trillion? They've never been used, and, even if they were, I would not be betting on eSafety winning in a courtroom. The agency's poor understanding of the laws that govern it was brought into focus by the recent Baumgarten decision, where eSafety made an informal takedown request without reference to any part of the Online Safety Act then, when challenged, declared that it had no case to answer as there was no official decision under the act to review.

A lawyers' picnic ensued, which eSafety clearly lost at great expense to the taxpayer. That was against an individual who had a post on X taken down. What hope would eSafety have in a battle against Meta or Google and their lawyers?

Very little, I would suggest. What is the other earth-shattering reform to the social media ban? Instead of the eSafety Commissioner only being able to seek information and documents from social media platforms and electronic services, the commissioner would now be able to ask anyone who has information relevant to a platform's interaction with the social media ban.

Yes, that's probably why the ban isn't working, because 85 per cent of kids under 16 are still on social media because the existing information-gathering powers aren't broad enough. These powers, as drafted, would allow the eSafety Commissioner to order parents and even children to produce documents. New amendments circulated last night would also empower eSafety to summon people and compel them to answer questions or risk imprisonment.

This is based on existing powers the eSafety Commissioner has in relation to other enforcement areas, but these ASIO-style powers seem disproportionate to the aim of stopping kids from having accounts on certain social media platforms. The late introduction of these amendments means that neither the Scrutiny of Bills Committee nor the Environment and Communications Legislation Committee have had a chance to consider and scrutinise them.

During the inquiry into this bill, it was revealed that artificial intelligence had been used in parts of the Age Assurance Technology Trial report, which informed the government's understanding of the tools available to assess users' ages. The Age Check Certification Scheme, who produced the relevant parts of the report, initially denied using AI at all. When they were presented with incontrovertible proof by the Guardian, they did a 180 and said, 'Oh, yes, we used AI, and we actually already disclosed it, since we accidentally left ChatGPT metadata in the report.' Not only did they lie about the use of AI; they lied to the government about erroneous citation in the report.

They said that the links worked at the time the report was handed over, but analysis of archived versions of those links shows that they have always pointed to incorrect references. How does the scheme respond to these allegations? By name-calling in industry magazines.

In an email quoted in biometricupdate.com, the chief executive, Tony Allen, had nothing to say about this misleading of multiple parties and instead asserted that the government was— (Time expired)