The already serious, chronic issue of bullying has now taken a new face. AI chatbots.
The traditional, almost stereotypical face of in person, peer to peer bullying now comes in a far more sinister presence unconfined by the physical proximity of schools and playgrounds, ready to spread harm 24/7.
This week, the federal government announced a national plan to combat school bullying, with the focus targeting AI chatbots that the Education Minister Jason Clare characterised as "supercharging" the problem.
The plan includes a 10 million investment that funds a national awareness campaign with new resources targeted at parents, teachers and communities.
However, the issue of the viability of such resources being effectively disseminated and made aware of in communities is being brought into question. The plan ideally “requires schools to respond to a bullying complaint or incident within two days,” yet the anonymised nature of AI being used for bullying as well as the overall reluctance of platforms to actively enforce restrictions has raised questions how this applies to this emerging threat.
Furthermore, the 2024 social media ban for under 16s have seen mixed results thus far, with some platforms restrictions being scrutinised due to entry barriers as low as a mere warning with no verification posing concern, and a lack of review on pre-existing teen accounts not being suspended or verified.
Minister Clare further warned of the emergence of AI bullying.
"I don't know if you guys have heard this before, I hadn't ... AI chatbots are now bullying kids. It's not kids bullying kids," he told the reporters.
"It's AI bullying kids, humiliating them, hurting them, telling them they're losers, telling them to kill themselves.
"I can't think of anything more terrifying than that, than AI telling kids to kill themselves. We've heard stories overseas of kids doing that. That's what we're up against." (Source: ABC)
Currently, one in four students between years four and nine have reported bullying every few weeks or more, according to the National Anti Bullying Rapid Review.
It’s commendable that the government has taken initiative to address this, but a more comprehensive approach is needed beyond mere distribution of resources. Student awareness of the risks and harms as well as strategies to combat such forms of bullying chronically lag behind.
The aftermath of being underequipped against such threats is already prevalent globally, and seeing its mark in Australia.
In October 2024, a lawsuit in the US alleged that an AI chatbot pushed a teenager to take his own life, and in September, American parents whose children took their own lives said they were encouraged to do so by AI chatbots during a US Senate hearing.
In August, the ABC reported that a teenager in Australia was encouraged to commit suicide, according to his youth counsellor. The 2025 Norton Cyber Safety Insights Report was released in July, and two in five Australian parents believe their kids are seeking AI for companionship. But do they know the risks associated with such forms of interaction?
Whether or not software and apps designed mainly by conglomerates in the U.S. are willing to actively collaborate with the national anti-bullying framework is also another issue.
"An app that's developed on the other side of the world can hurt a child here in Australia. And that's why we've gotta take this seriously".
The prospect of AI tools exploited to more serious crimes has simultaneously arisen, with AI-driven "nudify apps", which can be used to create non-consensual sexual material, were another source of deep concern especially if targeted against underage children. The complete lack of ethical considerations and safeguards in place, both by the user and the platform may very well exacerbate a sexual harassment crisis.
The eSafety Commissioner reported in June that digitally altered intimate images, including deepfakes, of people under 18 more than doubled in the past 18 months. 80 per cent of the targets were women. In early September, the federal government announced plans to restrict access to deepfake tools and specifically “nudify” apps. We foresee this to only move such tools underground to the dark web without removing the perpetrators of such crimes. Action in dealing with illicit misuse must come in conjunction with surface restrictions and bans.
We don’t think it’s too late for something to be done here. Ideally, legislative reform should advance to tackle the more serious sexual harassment aspects, but for now, the menace of AI in worsening bullying throughout schools is currently being tackled by overburdened teachers and parents who don’t have any comprehensive strategies and resources to thoroughly ensure child safety.