Allegra Spender (Wentworth, Independent) | Hansard source

Last night, in his annual threat assessment, the Director-General of ASIO, Mike Burgess, warned that Australia's rising terrorist threat is being fuelled by social media. His warning was stark. In November last year, he cautioned that platform algorithms can escalate a user from non-violent content to grotesque ISIS propaganda within a few clicks. The radicalisation he describes is happening here in Australia. As we all know, just a month later, on 14 December 2025, we saw Australia's largest terrorist attack in history—an indescribably awful attack of antisemitic terror in my community in Bondi. For years I have thought about what we can do to combat violent extremism. In the months since, that question has only grown more urgent.

In Burgess's February 2025 assessment, more than a year earlier, he noted that fewer than 17 per cent of the minors that ASIO had investigated were born offshore. Of those, the median age on arrival was 4½. These are our young people radicalised online, often in their own home. In his 2024 submission to the Senate inquiry into right-wing extremist movements, ASIO gave evidence that online platforms remain significant enablers of radicalisation and the spread of propaganda. This evidence is consistent. The same dynamic drives misogyny. In 2024, Burgess explained that just 10 minutes spent looking at incel material prompts the algorithm to recommend increasingly violent, misogynistic propaganda, including posts glorifying incel terrorists. We know all too well that incels have committed terrorist acts overseas.

Anyone with a social media account understands a pattern: algorithms drive division, isolate users and spread hateful content. They are not transparent. They shape our way of thinking without our knowledge, normalising extremism and hate. The government has acted to protect young people through the social media ban for under-16s, and I welcome it. But the ban is still not perfect and is not fully implemented. Still, the algorithms reach everyone of all ages and stages. A poll by Teach Us Consent found that 63 per cent of respondents knew someone radicalised by the algorithms. Teenage boys are now amongst the most likely perpetrators of child sexual abuse—including 17- and 18-year-olds, who fall outside the social media ban. Research from the Global Institute for Women's Leadership ranks Australia among the most misogynistic nations in the Western world. Reports of sexual assault sit at an all-time high.

Teach Us Consent's Fix our Feeds initiative offers a clear answer: put informed choice back into the hands of Australians. The European Union's Digital Service Act already requires major platforms to offer users an option to switch and opt out of personalised, algorithm driven feeds. The platforms complied. TikTok's 'For You' feed can show content popular in a region rather than content driven by past behaviour. On Instagram and Facebook, a fully non-personalized feed already exists. Under such an option, Australians could have the same experience. This sits well within the capabilities of these platforms. There is no question of that. We cannot leave users to the game of whack-a-mole, waiting for platforms to deplatform one creator after another. Beyond the algorithm, there are further harms we must confront.

I commend the government on beginning its work on a digital duty of care, which would shift online safety from a reactive, user supported system to a proactive, risk based model, legally requiring tech companies to anticipate and mitigate foreseeable harms with continuous safety-by-design standards across platforms, operating systems and connected devices. There is a long way to go. Part of that work is dispute resolution. Constituents and small businesses frequently come to my office in distress—businesses locked out of Instagram accounts they rely on to advertise and individuals shut out of their email—and they find it extraordinarily difficult to find any path of resolution. We have heard from small businesses plagued by fake Google reviews and even by fake accounts impersonating local owners to leave disparaging reviews on neighbourhood businesses. These are serious issues, and they strike at some of the largest disruptors of our social cohesion. It is crucial that this parliament deals with them dynamically, attentive to the detail and to the changing technology landscape.

Algorithms matter. Children deserve protection from the harms of social media, but so does everybody else. So it is time for us to be brave. It is time for us to fix our feeds. It is time for Australians to be able to permanently turn off their algorithms and choose when they are turned back on so that they have control over their social media. That's what I want for my kids as they grow up and what I want for my community, and it's what they want as well. Our businesses want a way to resolve issues with social media companies in an appropriate dispute resolution forum.

See this speech in context