Home Affairs Minister Tony Burke has cautioned that extremist groups may soon use artificial intelligence (AI) chatbots to groom young Australian men, underscoring the accelerating risks posed by online radicalisation. Speaking at the National Press Club on Thursday, Burke highlighted the speed with which individuals can now be radicalised, warning that what once took years could happen within days.
Burke expressed concern about young males engaging in online gaming and conversations with individuals they believe to be peers, who may in fact be AI-driven entities designed to facilitate radicalisation. He explained that once an individual reaches a certain level of radicalisation through interaction with AI, they could be handed over to a human recruiter.
In response, the government is developing a new information campaign targeted at parents, teachers, and coaches to help identify early signs of radicalisation, such as sudden changes in language or behaviour. The campaign will promote the Step Together helpline, initially established in New South Wales and expanded nationally since July 2025 with $6.8 million in funding over three years. Burke urged the public to use the helpline proactively, even if they are not certain that an individual is fully radicalised.
Burke’s remarks followed revelations that three male teenagers were involved in significant terrorism plots disrupted during the first half of 2026. These foiled attacks are among four “imminent” plots thwarted since the December 2025 Bondi massacre. However, only one incident has been publicly detailed, and authorities have not classified any of the disrupted plots as major terrorism threats in official police statements.
Additionally, Burke confirmed that the government is examining potential revisions to Australia’s national terrorism threat level system. The current level has remained at “probable” since August 2024, indicating a greater than 50 percent likelihood of an attack or its planning on Australian soil within the next 12 months. The next higher category, “expected,” requires concrete intelligence about a specific attack. Burke noted that the threat framework lacks nuance for escalating risk within the “probable” level, making it difficult to communicate incremental changes in threat severity.
The possible revision follows earlier comments from ASIO Director-General Mike Burgess, who has described the “probable” level as insufficient for conveying the full picture of the threat environment. Burgess said he is in discussions with Home Affairs Secretary Stephanie Foster about overhauling the threat level system to provide more detailed assessments.
The government’s cautious approach reflects ongoing efforts to address the evolving tactics employed by extremist actors, particularly with the advent of AI technologies that could intensify recruitment and radicalisation efforts online.