A practical Australian guide to AI companions, chatbots, privacy, deepfakes, emotional dependence, and the risks of using AI for support, advice, or therapy-adjacent care.
AI is no longer a specialist topic. It is part of ordinary online life.
Young people use AI for schoolwork, entertainment, emotional support, advice, roleplay, image generation, and sometimes for things that look uncomfortably like therapy.
This article is one branch of the wider Online Safety in Australia pillar guide. The purpose here is to separate the useful parts of AI from the risky assumptions people are increasingly making about it.
The obvious risk is inaccuracy. That matters.
But many of the bigger risks are relational and privacy-based:
An AI tool can sound calm, empathic, and certain while still being wrong, manipulative, or unaccountable.
Some AI products are marketed less like tools and more like:
That matters because the design goal is often attachment.
For some young people, especially those who are lonely, distressed, or socially isolated, that can make the tool feel safer than it really is.
The problem is not only the content of what it says. The problem is that the relationship can become sticky, private, and persuasive.
Practical risks include:
Deepfakes sit inside this same wider ecosystem. If your main concern is fake explicit imagery or image threats, the best companion article is Sextortion, Deepfakes, and Image-Based Abuse in Australia.
Some AI tools now present themselves as:
The danger is not that every tool is useless. It is that people start to confuse emotional tone with accountable care.
For clinicians, the key issue is responsibility. Convenience does not replace:
For families, the simplest version is:
AI can be useful. It is not a therapist, not a crisis service, and not a trusted adult.
These questions matter for clinicians too, especially when AI tools are used for notes, summaries, or client-facing workflows.
AI safety is not only an individual-use issue.
It also intersects with:
That is why AI safety belongs inside the broader online safety conversation rather than sitting off to the side as a separate "tech topic".
The safest way to think about AI is not as magic and not as evil.
It is a powerful, persuasive, increasingly intimate technology that can be useful in narrow ways and harmful in others.
For children, families, and clinicians, the core task is the same:
Share your thoughts and experiences with this resource.
Sign in to leave a comment
Move from strategy into implementation with templates, handouts, and psychoeducation tools already live on the marketplace.
Publish clinician-grade templates, build trust signals, and start growing an evergreen library under your own brand.
A practical Australian guide to the under-16 social media rules, what platforms are age-restricted, what children can still use, where parents get confused, and what schools and clinicians should say about it.
A practical Australian guide to gaming spending risks, including currency distancing, loot boxes, battle passes, skins, gifting, status pressure, and why modern games can feel more like platforms than toys.
A practical Australian guide to parental controls, monitoring software, Screen Time, Family Link, console settings, purchase controls, and how to use oversight without turning family online safety into a secrecy war.
